Random selection in politics

Lyn Carson and Brian Martin

Published in 1999 by Praeger Publishers, Westport, CT Available for purchase from Praeger

This is the text submitted to Praeger in February 1999. It differs from the published version in minor ways, including formatting, copy-editing changes, page numbering (100 pages instead of 161) and omission of the index. This version prepared August 2008

Contents 1. Introduction 1 2. Random selection in decision making 10 3. Direct democracy 26 4. Citizen participation without random selection 36 5. Citizen participation with random selection: the early days 43 6. Citizen participation with random selection: yesterday, today, and tomorrow 52 7. Sortition futures 65 8. Strategies 76 Appendix: Examples of citizen participation 84 Bibliography 93

Acknowledgments Brownlea, John Burnheim, Ned Crosby, Many people we’ve talked to find the Jim Dator, Mary Lane, Marcus Schmidt, idea of random selection in politics and Stuart White. Their input was unnatural and unwelcome. This didn’t valuable even when we decided on an deter us! Fortunately, there were a few alternative approach. Helpful comments enthusiasts who got and kept us going. were also received from Ted Becker, Alan Davies, Fred Emery, and Merrelyn Stephen Healy, Lars Klüver, and Ken Emery provided the original inspiration Russell. Others who provided many years ago as well as ongoing information are acknowledged in the conversations. text. The process of obtaining comments Ted Becker encouraged us to write has been stimulating because not all this book. On drafts, we received readers go with us all the way on extensive comments from Arthur randomness.

1 Introduction

Government by elected representatives is to be involved in decision making is capable taught in schools and presented in the media as of making a useful contribution, and that the the natural way of doing things. Powerfully fairest way to ensure everyone has such an legitimized by the ideas of mandate and merit, opportunity is to give them an equal chance to representatives elected under this system be involved. Random selection worked in consider that the electorate has given them a ancient Athens. It works today to select juries mandate to govern, while bureaucrats consider and has proved, through many practical that merit and expertise justify their role in a experiments, that it can work well to deal with powerful decision-making elite. Representa- policy issues. tive government obviously is a great Random selection can be used to promote improvement over previous systems of rule both small-scale and large-scale political such as feudalism, autocracy, and dictatorship, participation, from a tiny exercise in street but nevertheless it is a system of rule in which improvement to a national electoral system. citizens have relatively little impact on a day- Like election, it needs to be used sensibly, to-day basis. with appropriate controls to ensure best Representative government has its limita- operation. tions. It concentrates power into a parliament There are various terms that can be used to or congress and the elected representatives can refer to random selection in decision making. become vulnerable to vested interests. The Typically we use the phrase “random selec- voters are given responsibility only for opinion tion,” sometimes abbreviating it to “random- formation, not decision making, and those ness.” An alternative is to refer to a “lottery” representatives who make the decisions have or the “lot system.” The latter is standard in low accountability. These and other problems discussing ancient Athenian democracy. are inevitable in representative government Finally there is the term “sortition,” which because it is a system in which a small number means the act of casting lots, which can be of people — politicians and high-level bureau- used to refer to choosing decision makers by crats — have a great deal of power which can lot, lottery, or random selection. be exercised to serve powerful interests, including their own self interests. OVERVIEW Most people attribute problems with repre- sentative government to individual politicians The chapters that follow show how random and specific policies. A standard assumption is selection can be used to create a more partici- that if only the right people could be elected patory democracy. Our general framework is and the correct policies implemented, then to first introduce various uses of randomness everything would be okay. But the problems in decision making (Chapter 2), then deal with go much deeper. examples and methods of participation in We want to step aside from a belief in the decision making without random selection ideas of mandate and merit as rationales for (Chapters 3 and 4), discuss current experience governance, since they are used to stymie with decision making involving random efforts to foster greater citizen participation. selection (Chapters 5 and 6), give an outline of We suggest instead a different foundation for possible political futures involving greater use fostering participation and diffusing power: of sortition (Chapter 7), and conclude with random selection. some ideas about strategies for moving The assumption behind random selection in towards greater citizen participation through politics is that just about anyone who wishes random selection (Chapter 8). 2 Random selection in politics

Chapter 2 sets the scene by illustrating the making in planning and policy making looks role of randomness as an explicit factor in good. decision making. It looks broadly at random Chapter 7 puts the case for random selec- selection in our everyday lives, for example at tion as a means of replacing representative the way gamblers toss a coin in the interests of government by direct citizen control. In fairness and the way in which Dutch universi- “demarchy,” groups of local citizens, ties choose between student applicants. randomly chosen from volunteers, deal with Random selection is shown to be a fair and policy on different functions such as industry, useful method for making decisions. It is, of education, and entertainment. Demarchy uses course, fundamental to those legal systems random selection to overcome the central that rely on it for selecting juries. The chapter dilemma of direct democracy, that not every- includes a description of the lot system in one has the time or interest to be involved in ancient Athens. making decisions about every topic. We Chapter 3 considers the possibilities pre- conclude in Chapter 8 with a discussion of sented by direct democracy, in which groups strategy for promoting random selection in of people run their lives directly without rul- politics, covering likely opponents, likely ers, elected or otherwise. Direct democracy is supporters, and opportunities for introducing evident in a number of historical and revolu- random selection. tionary events, through self-governing bodies Our aim is neither to undertake a compre- such as communes, soviets, councils, or hensive critique of the present system nor a committees, as well as in experiences in detailed examination of the literature. Rather workers’ control and community self- we emphasize innovative experiments and management. Other methods of direct possible applications of random selection, democracy include consensus, initiative, and aiming to provide clear descriptions and to referendum. We note limitations as well as raise ideas and questions while avoiding advantages of direct democracy. ponderous academic apparatus. Chapter 4 surveys a number of consultative The ideas in this book are rational but will mechanisms that have been used in technology not be popular with the elites whose power assessment, urban planning, and service and position will be threatened. Random delivery. Various consultative methods are selection undermines the claims to privilege discussed but they are ones that are currently based on appeals to merit and electoral devoid of a random selection component. mandate. Random selection may not even be Inevitably, these consultative options tend to popular with some of those within alternative replicate the corporate model of boards of movements, because it potentially threatens directors or representative governments and to their privileged positions. But these ideas are reproduce the very hierarchies they wish to at the heart of democracy and were the basis replace. They do, however, have potential to of the first political activities which stirred be transformed into fairer methods through the democracy into life in the polis of ancient use of random selection. Athens. This yearning for participation that is Chapters 5 and 6 offer a way out of this fair and inclusive strikes a chord for many. For mirroring of unsatisfactory representative some it has been prompted by a hatred of methods. Chapter 5 looks at the early days of government, for others by a love of modern citizen participation with random democracy. selection, particularly those methods that were Random selection should not be considered born in the 1970s and continue today — in isolation. Yes, it is a significant tool to citizens panels and planning cells. Chapter 6 transform politics. On its own it can increase continues the story, covering a number of fairness in decision making. Integrated with fascinating case studies from various coun- deliberation and consensus building it can tries. The future of participation in decision become a powerful means to achieve social justice and genuine democracy. Introduction 3

In the remainder of this chapter we outline government is quite young, less than a some problems with representative govern- thousand years old, and that voting by a ment arising from the nature of the party substantial proportion of the population is only system, the effect of power on leaders, the a couple of centuries old. Despite the youth- effect of electoral politics on citizens, and the fulness of representative government as it is existence of bureaucracy. It is useful to now practiced, it remains largely unchallenged understand the dynamics and shortcomings of as the established, accepted norm. This is representative government in order to despite its many flaws. Here we outline some understand why and how random selection can of the problems under the headings of the provide a useful means of reform as well as nature of the party system, the effect of power the basis for a full-scale alternative. on leaders, the effect of electoral politics on citizens, and the existence of bureaucracy. PROBLEMS WITH REPRESENTATIVE GOVERNMENT The Party System

Western representative governments seem at The biggest problem with political parties is first glance to be extremely fair — govern- that elites within them develop a vested ment of the people, by the people, for the interest in their own power. Party elites act to people. Everyone has an opportunity to put serve the party and themselves, often at the themselves forward as a candidate for election. expense of the public interest. The tendency of Every adult citizen has an opportunity to vote party organizations — including revolutionary in elections for people who will represent their ones — to become less participatory and more interests. If representatives don’t perform, they oligarchical was expounded by Robert can be voted out of office next time around. Michels nearly a century ago, and little has What could be fairer? Unfortunately, the changed since then (Michels [1915] 1959). fairness of this representative system does not In party-based representative government, withstand close scrutiny. A “fair” competitive voters get to choose between different candi- process is not necessarily a “fair” system in dates on the ballot, but many or most of the outcomes. Most people are excluded from candidates are attached to political parties. A direct decision making and those who become candidate who is a party member is tied, representatives are far from representative of tightly or loosely, to the party and its policies. the general population. Gone are the days when genuine representa- Note that there is often a big difference tives of constituencies were chosen. Voters are between “being a representative” — that is, left to choose a color or flavor, a brand, a being elected to a post on behalf of a constitu- package of policy products. Voters may ency — and “being representative” — namely, approve of some of the policies and not others, being a typical member of a community. We but they can vote only for the entire package. refer to the latter as “representativeness.” There are many different systems operating In systems of government that are called even within the broad category called Western “representative” or “electoral,” representatives representative government. In Britain and are elected via the ballot box with voters being Australia, for example, there is quite rigid given the option to vote, as they are in the U.S. party control of elected representatives. If or Britain. Voting in Australia and Belgium is members of parliament vote against the party compulsory, which means that citizens are line on any issue, they are likely to be required to attend a polling booth. It is ostracized, expelled, or not endorsed at the assumed that this is how it should be, that this next election. This, by the way, has overtones is the basis of democracy — that candidates of the “democratic centralism” of a communist stand for election and citizens who are party, which means that no member of the prepared to vote, have the right to do so. It is inner core can deviate from the party line. worth remembering that representative 4 Random selection in politics

In the United States a somewhat looser Random selection, as an alternative to system exists, with party members more freely election, undermines the party system. Parties crossing the floor of Congress. Rather than would no longer be able to control who is being rigidly locked into party loyalty, nominated and vested interests would have a congressional representatives are accused of harder time buying support, since randomly being heavily influenced by special interests, selected decision makers are beholden to no typically corporations, that provide campaign one for their position and, without an electoral donations (Stern 1988). In contemporary mandate, have no basis for being selected marketing culture, parties, politicians, and again. policies are advertised and “sold” just like The founders of modern systems of repre- detergent, movies, or any other commodity, in sentative government in Britain, France, and what can be called “promotional politics” the US recognized the oligarchical or “aristo- (Wernick 1991: 124-153). cratic” tendencies of electoral systems: the Some European and other countries have an tendency of representatives to be superior to electoral system of multi-member electorates the electorate. Representative government was or proportional representation. Proportional seen by its founders, and by leading political representation gives candidates from smaller commentators at the time, as qualitatively less parties a greater chance of being elected. democratic than the lot system. Although Inevitably the larger parties are able to create a representative government has changed in the more impressive public image because of their past couple of centuries, its characteristic access to more funds, so they still remain feature of being a “democratic aristocracy” has advantaged. remained (Manin 1997). Politicians are divided into two camps: in a parliament there are backbenchers and the The Effect of Electoral Politics on Politicians executive, with the latter making all the significant decisions; in congress the same In order to be even remotely effective, separation exists between those who chair or politicians need to do an enormous amount of are members of powerful committees and work. The range of topics on which they are those who are not. The lowly representative is expected to vote is extremely diverse and it reduced to voting fodder in parliament or would be impossible to possess expertise in all congress. areas. As well as learning about the issues, With all of the electoral systems that have politicians must spend time dealing with party been mentioned, the elected representatives pressure as well as pressure from lobbyists on become full-time politicians. All of the people behalf of powerful corporations and other who stand for election must be prepared to special interest groups, not to mention requests enter into a culture that sees a separation from constituents. Not least is the need to between citizens and elected officials. The attend meetings, give speeches, attend open- notion of community service is increasingly ings, and many other activities necessary to being replaced by a well-paid career. maintain a profile and continue to be elected. It is a typical feature of every representative So far, we have referred to national or state government that elected representatives are politicians. Representatives at a local level, for not typical members of the community: example, in local government, do not receive inevitably, most of them have greater wealth, high salaries, are less likely to be involved in status, or perceived talents than most of those political parties, and are subject to less who vote for them. This is especially notice- pressure. Their workload is still high and they able in labor parties, where few representa- are also likely to become enculturated into tives have ever spent much time as typical their elite decision-making bodies just like workers. How many hairdressers, mechanics, their state or national colleagues. One of us or sales assistants are ever elected to office? (Carson) speaks from personal experience from a period in local government when the Introduction 5

role of “representative” on many occasions an end. There have been occasional exceptions became a stronger influence than that of being to the adversarial two-party model. Denmark, a “community activist” or even “community for example, had minority governments for all member”. Elitism can subtly permeate the but a few years in the 1900s; its parliamentary psyche. decision makers learnt to work with compro- Elected representatives come in for some mise and consensus (see Chapter 4). rough treatment at the hands of their constitu- Social movements and community activists ents for “faulty” decision making and this find themselves sucked into this swamp of presumably affects their decisions. Carson adversarial politics. Like the governments they conducted interviews regarding the conse- condemn, they are vulnerable to replicating quences of decision making with local hierarchies and placing considerable power government representatives from one rural into the hands of a few. As lobbyists they learn community. These councillors reported that to play the game, making deals and trying to they had been ostracized, abused, and received extract assurances from political parties that obscene phone calls and death threats because won’t always deliver when the political winds constituents disagreed with their decisions change. Social movements find themselves (Carson 1996). appointing leaders who share the charismatic Critics of a process or an issue can often be media personalities of the politicians against much more tenacious than those defending or whom they do battle. Charisma is not neces- supporting it. For activists who later become sarily a quality that is consistent with wise public officials, the transition from “critic” to decision making. “leader” or “policy maker” can be a most Experiments provide empirical support for difficult and confusing one to make. Czecho- Lord Acton’s aphorism that “power tends to slovakian dissidents who became the govern- corrupt and absolute power corrupts abso- ment overnight found the transition from wise lutely” (Kipnis 1981, 1990). Politicians are critic to wise leader remarkably difficult highly susceptible to the corruptions of power. (Atlee 1991). The enthusiasm of newly elected Random selection would remove much of the representatives can be subtly and gradually power and hence reduce the possibility of paralyzed by the system (Carson 1996). corruption. Nelson Mandela noted “In some ways it is easier to be a dissident, for then one is without The Effect of Electoral Politics on Citizens responsibility. As a member of the Executive, I had to weigh arguments and make decisions Elections empower politicians a lot more than and expect to be criticized by rebels like they do voters. What little power the voter has myself ” (Mandela 1995: 135). exists primarily on election day. Though a Representatives in Western governments right of recall exists in some parts of the U.S., function within an adversarial model that its enactment is a rarity. Therefore, should permeates local, state, and national levels of politicians prove to be bone lazy or ineffec- governance. All spheres of government tual, they remain. There is no real obligation to operate in a culture of conflict. This prevailing meet any promises. Indeed the community culture can become a daily ritual of goodies increasingly expects politicians to lie and polls and baddies, reinforced by the media that too indicate that citizens believe they are not to be often gleefully reports or even manufactures trusted. Comparative U.S. polls, for example, conflict. The energy expended on electioneer- indicate that whereas three out of four of those ing fanfare and the ongoing slandering of asked in the late 1950s trusted the federal individual politicians might be better spent on government, by the 1990s just one out of four getting down to the business of decision trusted it, a dramatic turnaround in public making — quietly, deliberatively, consensu- opinion (Orren 1997: 80). There is no way to ally. True, dissent is a healthy component of sue a politician for false advertising, for not democracy but presumably only as a means to delivering election promises. Conversely, 6 Random selection in politics

there is little support for politicians from a and electorates. Although voting gives some disaffected community. Instead of supporting power to citizens, it can increase the power of leaders to lead better, citizens reduce them to governments by giving them much greater stereotypes or objects of scorn. One small legitimacy (Ginsberg 1982; Manin 1997). exception to this is the Heart Politics Further, elections serve to control otherwise movement, now located in a number of unmanageable political activity. Voters learn countries, particularly Australia, New Zealand, the limits of political activity, namely to elect and the U.S. The movement is directed leaders, not to determine policy. towards a politics of connection rather than There are other problems associated with confrontation and of finding ways in which having representatives speaking on behalf of leadership might be supported in order to be others (Martin 1988; Morgan 1988). The reformed (Peavey 1984). whole idea of representation or delegation has It is evident that the level of acceptable come in for criticism and has even been political participation has increased during the rejected in some quarters. Speaking for others past two hundred years of representative has been seen as “arrogant, vain, unethical, government. However this level of participa- and politically illegitimate” because of the tion has become quite a routine activity and is difficulty of representatives transcending their mainly limited to voting (Martin 1995). own social identities and the importance of Benjamin Ginsberg has argued that the allowing the oppressed to speak for them- expansion of suffrage resulted in a reduced selves (Alcoff 1991: 6). There will always be level of radical direct action (Ginsberg 1982, times when speaking for others is essential, for 1986). Ginsberg also makes a number of example on behalf of those without a voice, salient points about the way in which the such as young children and people with public can be manipulated in the formation of profound intellectual disabilities. There will its opinions at the same time as the state always be occasions when others will prefer to makes a show of ruling according to the very appoint a delegate. opinions that have been cultivated (Ginsberg However, elected representatives are not 1986: 223-226). This may be why some delegates: they do not have to do what citizens people have the unnerving feeling that want. In other words, voting is making a politicians and the media seem to have a choice between candidates, not making different agenda from their own. political decisions directly. Voting allows little Lukes addresses this idea of agenda setting room “for texture or nuance and does not when describing power as three-dimensional. create space for transformation or change” Yes, agenda setting is done by those who hold (Phillips 1995: 41). Even though opportunities power, according to Lukes, but non-decision for voting have increased, so has the power of making is just as important as decision the state. The result can be a sense of isolation making. Demands for change can be suffo- and powerlessness in an increasingly complex cated before they have a chance to be aired but system of social administration. Popular beyond that, potential issues might be neither control seems ever more remote, though some consciously chosen nor the result of particular retain the fantasy that it exists. individual’s choices. Instead a more subtle Random selection can help overcome the socialization can occur via social forces or learned powerlessness of many citizens, who institutional practices that result in certain would not dream of putting themselves interests not being expressed or even known forward for election, even if they could afford (Lukes 1974; see also Gaventa 1980). to. Just as juries empower citizens in the legal The expansion of suffrage is typically system, so random selection can do so in many presented as a triumph over privilege. Despite other areas of decision making. opposition from the propertied class, workers gained the vote. Women, too, were awarded the vote despite male-dominated governments Introduction 7

BUREAUCRACIES given to authorities about how to do this. The result can be a plethora of public notices in Many of the decisions that affect people every newspapers advertising proposed development day — for example in employment, housing, applications and this in turn arouses predict- and education — are made by bureaucrats, able hostility from a disgruntled community who are even less subject to popular control that is weary of pseudo-democratic processes. than politicians (Phillips 1995: 39). Citizens Bureaucracies are also workplaces. Some have little hope of changing bureaucracies researchers argue that decision-making whether they are government departments, processes in the workplace affect whether and corporations, labor unions, or churches. how people become politically involved Nevertheless, many people working within (Pateman 1970). If there is no experience of bureaucracies understand the need for citizen affecting decisions at work, citizens will be participation. Participation is seen as a way to less inclined to be politically active in the improve service delivery, a strategy to avoid wider community. Dahl (1985: 111-135) costly litigation when a project goes wrong, or argues that if democracy can be justified for a means to enable bureaucrats to gather the governing a state it is equally justified in a sort of information that is essential before they firm. can confidently make good decisions. This Can society operate without bureaucracy? participation can come too late, can be Not the way it’s presently organized. But it is tokenistic, or even manipulative — just a possible to imagine running a complex society shallow form of public relations. At its best, without the sorts of bureaucracies that exist however, there are well-meaning bureaucrats now, using randomly selected groups of who are often very unsure about how citizen decision makers, as described in Chapter 7. participation should occur. They can also feel quite fearful about involving the general PROBLEMS WITH PARTICIPATION public. Bureaucrats have experienced enough out-of-control public meetings to be justifiably If citizens become policy makers, does that wary. mean that visionaries will not emerge? Martin Political theorists now speak of an era of Luther King, Jr., Nelson Mandela, Vaclav post bureaucracy in which more consultative Havel, or Aung San Suu Kyi are unlikely to mechanisms must be instituted so that citizens have emerged as formal leaders through a can regain a degree of influence (Laffin & process of random selection. If decision- Painter 1995). Planning, service provision, and making duties are rotated, such bright stars evaluation are areas in which citizen partici- might not have a chance to shine. Interest- pation would seem to be essential. Every time ingly, though, when we think about visionar- a local government makes a decision to rezone ies, it turns out that they obtained their land or approve a major development, reputations without or before being elected to community members are affected. Each time a office. The systems that we outline are more regional or national government department likely to provide space for visionaries in makes a decision about expanding, withdraw- public life. ing, or in any way altering health, education, Participation can be subject to manipula- or other community services, the citizen or tion. It can be abused by those who would consumer is affected. wish to use citizen involvement as a public Most local government planning policy relations exercise or a means to engineer an development in Australia is accompanied by outcome (Arnstein 1969). If participation only citizen participation; in the U.S. this is more leads to “citizens proudly and cheerfully variable. Some legislation (for example the [waving] their own chains” (Ginsberg 1986: Local Government Act in New South Wales, 232), as it often does now, the cause would Australia) specifically requires community have been lost. consultation. Unfortunately little advice is 8 Random selection in politics

Public participation or participatory Apathy is socially produced rather than innate democracy is not without its critics. Some (Eliasoph 1998). It may be more useful to would say that it takes too much time, that the liken the public’s unwillingness to an unused average person does not want to be bothered muscle; because the public is not consulted, its with involvement in every issue (Mansbridge willingness to do so has atrophied; with 1973). Citizens have jobs to do and friends sufficient exercise this can change (Carson and families to connect with. We are mindful 1996: 161). With greater access to decision of this point, which is addressed (for example making the community is more likely to in Chapters 5 and 6) as we traverse the participate and is educated by the process grounds of voluntary participation and the itself. This in turns leads to increased aware- extent to which this leads to the swamp of ness, less indifference, and an avoidance of the self-selection. The Swiss experience certainly “over consulted” feeling which comes from stands in sharp contrast to the belief that having no influence on decisions. In other citizens will tire of too much participation. words, people “learn to participate by partici- The levels of citizen participation in Switzer- pating” (Pateman 1970: 105). Carol Gould land remain remarkably high despite the sums it up as follows: “the fact that people fail demands of regular referenda and citizen to vote or to participate in democratic councils. processes is not the result of their unwilling- Any participation that is meeting-based ness to be active or to exercise their powers tends to favor men. Feminists have long but rather it is because they believe their argued that greater political involvement is activity in these instances would be futile … skewed towards men — that it is not gender They regard such procedures as a sham or neutral (Bussemaker and Voet 1998). This has merely ritual … where people believe their been our experience and other practitioners’ participation is effective, they are more likely experience and, again, is discussed in the to participate” (Gould 1988: 296). In order to chapters that follow. Electronic democracy is create a “strong democracy” (Barber 1984), often raised as a means to counteract sexual participation needs to be institutionalized. inequalities. It is necessary to be aware, however, of the shortcomings of some direct PARTICIPATION IN LATE- techniques, especially those that favor “under- CAPITALIST SOCIETIES comprehension” and simplistic reductions of complex arguments into a yes/no response. Citizen participation in Western representative Any reform to decision-making processes governments is hampered not just by powerful must take into account the needs of all politicians or unwieldy bureaucracies or citizens, regardless of class, sex, ethnicity, or national governments. The late twentieth ability. century heralded a period of late-capitalism Apathy is always raised as a problem when that defies national borders. Powerful owners discussing how to increase levels of citizen and executives are not elected. Their existence involvement. In the U.S. the low voter turnout is validated through reference to the need for is often attributed to this phenomenon. stockholder profit. Corporations are seemingly Citizens are dismissed as “apathetic”, wishing beyond restraint. Cheap labor is sought in to leave government to politicians and bureau- developing countries, tax havens protect the crats while reserving the right to criticize. swelling financial coffers of multinational There is some doubt, in this common view, corporations, currency speculation is beyond that people would actually want their the taxation systems of states. It could be democracy invigorated if it required any effort argued that national sovereignty is becoming on their part. Chapters 5 and 6 provide an anachronism and the power of politicians evidence to the contrary. and bureaucrats is mouse-like when compared However, politicians and commentators to the roar of the lions of capitalist enterprises seem to be locked into this belief in apathy. Introduction 9

in an increasingly uncontrolled global market exercise. It is possible to ignore Margaret place. Thatcher’s famous statement in the 1980s that Thankfully, a few of the globally affluent there is “no such thing as society”, and to are beginning to express outrage at their own begin to create civil societies based on legal, but morally questionable, actions and increasing social capital. The following the impacts these actions are having on chapters endeavor to breathe life into a method weakened nation states (Soros 1998). World of decision making which has always been citizens are creating issues-based movements available but has been suffocated by methods to react to these global forces — the campaign which are unfair, are stifling genuine debate, against the Multilateral Agreement on Invest- and inevitably are leading to poor decisions. ment in the late 1990s is just one example. For democracy (in the government, the Citizen participation is taking different forms organization, or the community) to be strong it and the necessity for citizen involvement in must contain the essential element of citizen decision making has never seemed more participation, not just participation by a self- pressing. selected few but participation by ordinary people who rightly can determine their own CONCLUSION futures. Given the difficulty of involving everyone in such a deliberative process, we Democracy, this ideal which lives in the hearts argue that random selection is an ideal means of so many people, is best seen as a process, by which a cross section of the population can not as a system or structure of government but be involved. as an ideal towards which communities and individuals strive: a verb (“democratize”), rather than a noun. There are many definitions of this ideal but there is general agreement that it indicates a sharing of power, joint setting of agendas, sharing information, and decision making. It is claimed that participation at the local level is a means to “learn democracy” because it allows residents to learn the “rudiments of self-government within a smaller unit” (Pateman 1970: 38). This participation need not be confined to the local or to specific interests. Decision-making approaches such as those in the following chapters are process-oriented and sensitive to different contexts. Such approaches are often not popular at a time when there is greater emphasis on the individ- ual and a hunger for quick-fix solutions. The alternative decision-making processes that we outline are as relevant for those working at the grassroots, building communities, as they are for those in government who genuinely are seeking better ways of involving citizens in the decisions that affect them. It is possible to move beyond the cynical public realm of politics which is so often “confined to specta- cles and acclamation” (Habermas 1971: 75) where politics is seen as a public relations 2 Random selection in decision making

The use of random selection in decision was extremely popular, especially from the making is a way of being fair and being seen mid 1800s to the mid 1900s. to be fair. It is especially valuable when other Imagine an alternative: each coin is held in methods leave open a suspicion of bias or one hand of a “coinholder,” behind her back. inappropriate conflict of interest. After the betting, she puts her hands out, In everyday usage, the word “random” palms upwards, and reveals heads or tails. If means haphazard, aimless, or without a the coinholder is totally honest, everything is definite purpose. However, we use the word in fine. But some players may suspect that the its statistical sense, in which a random sample coinholder, noting how the betting is is one in which each member of a population proceeding, may be changing the coins. It has an equal probability of being selected. In doesn’t matter whether the coinholder is other words, random selection means equal totally honest or not: there is no way of chances of being selected. There can be a very convincing everyone that she is honest. definite purpose behind the use of random- Tossing the coins ensures that the result is ness, though the actual selection made is a seen to be fair, because it cannot be influenced matter of chance. by anyone involved. The coins, of course, Using random selection requires quite a bit have to be checked to make sure that they are of judgment, just like any other method of “fair” or “true,” namely having an equal decision making. It does not automatically chance of giving a head or tail on each toss. produce fairness. At what point in the process Likewise, any other possible human influence should random selection be applied? How are on the result needs to be removed. Two-up options selected? When are people persuaded rules require that the coins must reach a that randomness is fair? What must be done to certain height and spin suitably. The coins are ensure fairness? Here we look at four major specially colored to prevent the spinner ways in which random selection can be surreptitiously introducing double-headed deployed: to make decisions directly, as a coins. When all this is done properly, losers strategy to make decisions, to judge opinions can only blame bad luck and not bias or and to choose decision makers. Rather than conspiracy. present abstract arguments (Broome 1984), we Other games of pure chance, such as lotter- mainly raise ideas through various examples. ies and roulette, are analogous to two-up. The We also look at evidence about perceptions of result must be seen to be due purely to random the fairness of random selection. processes that cannot be predicted or con- trolled by anyone. (In practice, there are RANDOM SELECTION TO MAKE various scams to cheat the system, which DECISIONS means producing a non-random result.) In many lotteries and prize draws, the rules Games specify that employees and family members of the company sponsoring the enterprise cannot In many games of chance, randomness is seen enter. Why not? It is to ensure that the result is as central to fairness. Consider the Australian seen to be fair in outcome as well as process. game of two-up: two coins are tossed in the air If the lottery director’s daughter won the top and players bet on whether there will be two prize, many people would suspect cheating heads or two tails. When the result is one head even if the result was pure luck. To avoid any and one tail, the coins are tossed again. Simple appearance of special treatment for insiders, enough. Two-up, though illegal in Australia, only “outsiders” are allowed to play. This Random selection in decision making 11

suggests that being seen to be fair can be as team and a greater chance of a win for the important as actually being fair. satisfaction of local spectators. This is allow- When the outcome is purely random, games able because teams play equal numbers of can be, and be seen to be, “clean.” Adding just games as the visitors and as the home team. In a touch of nonrandomness messes everything cricket, though, batting last is not always an up. Consider horse racing. Much depends on advantage, depending on the state of the known factors, such as the speed of the horse, wicket and other factors. A coin toss is used to its weight, rider, and start position. But of decide which cricket captain gets to choose course not everything is known, since other- whether to bat or field first. wise the winner could be predicted with 100% Games, whether sports or games of chance, accuracy and there would be no point in are artificial situations designed and used for a holding the race at all. Chance factors play a variety of purposes, including amusement, significant role, such as a horse or rider being profit, and personal achievement. In some off color on the day, getting caught behind sports, such as professional wrestling, there is other horses, or putting a foot wrong. Because no pretense that the event is fair. But fairness both random and nonrandom factors are is commonly a key element in games, which is involved, the way is opened for manipulation. why rules are enforced and cheating is Some interventions are seen as legitimate, condemned. Various methods are used to such as training the horse. Others, such as maintain fairness, including referees, scruti- drugging the horse or holding it back during neers, and drug testing — and randomness. the race, are seen as illegitimate, because they can be used by insiders to make money at the Among Friends expense of other gamblers. Hence it should not be surprising that horse racing has been Imagine that you are one of a small group of subject to periodic cheating scandals, and no friends who need to make a choice. When doubt much cheating is never detected. The might you use random selection? Suppose same applies to sports such as soccer, baseball, someone gives your group a reasonable sum of and boxing where players affect the outcome money, to distribute as you please. It might be and thus have an incentive to throw the game that Mary needs the money more than you do. or change the result by a few points. However, unequal divisions can cause Sport is an arena where fairness is abso- problems. Exactly how much should Mary lutely crucial. The rules of sports establish a receive compared to the rest of you? You may micro-world in which wider social and politi- feel resentful that Mary is getting so much, or cal influences are set aside. All that is Mary may feel obligated as a result of her supposed to matter is performance within the extra amount, or both. Resentment and a sense boundaries of the event. The key reason why of obligation can cause tensions in the group. drugs are banned in many sports is that a drug- This can be avoided by dividing the money free competition — in which no contestants equally. Equal allocations often are seen as have an artificial advantage — seems fair. fair even though need is unequal. This view is implicit in much writing about So far, so good. But money is relatively drugs in sport (e.g., Mottram 1988; Wadler easy to distribute because it is divisible. What and Hainline 1989). In competitive sport, the if your group receives one free ticket to a live rules are designed to ensure that the better performance by your favorite artist? The performer wins. Cheating that is exposed may performance is sold out and this may be the be penalized severely. last time the star will ever perform. If you are If there is a systematic advantage to one the only one who likes the star, the solution is side in a particular contest, then it is expected easy — you get the ticket. But what if all of that this will be compensated at another time. you adore the star and would value the ticket In baseball, the visitors bat first and the home more than any amount of money? In this team last, giving an advantage to the home situation, using a random process such as 12 Random selection in politics

flipping a coin or tossing some dice may seem pain relief than the ones getting the placebo. fair to everyone. Everyone has an equal The next complication is deciding who gets chance. the drug and who gets the placebo. If the If, on the other hand, you expect to receive researcher decides, then there might be some a series of free tickets over the coming bias because a certain type of person is chosen months, it often seems fair to alternate: one to get the drug. Furthermore, if the researcher person goes one time, another person the next. knows who is getting the drug, the researcher But who goes first? Flipping a coin is a good might give subtle unconscious indications to method if there are no other ways to decide. the volunteers, again biasing the result. Among friends, random selection thus can To avoid bias and placebo effects, then, the be a useful method of decision making when researcher may choose to run a randomized other methods run into problems. Often double-blind trial. Volunteers are chosen decisions can be reached by mutual agree- according to strict criteria and put into two (or ment, for example unequal divisions based on more) groups that are matched to be as similar need or merit, equal divisions, or having as possible. The choice of who gets the drug different people have their choice at different and who gets the placebo is made randomly, times. Random selection is helpful when none and neither the volunteers nor the researcher of these methods works or when they cause knows which is which during the trial. (In a tensions in the group. “blind” trial the subjects don’t know who is A final situation is when no one really cares getting what; in a double-blind trial the about which choice is made. Tossing a coin researcher doesn’t know either.) This is the about which restaurant to go to can help basic idea behind much clinical testing and overcome a deadlock when no one wants other research. responsibility for making a choice. It can also Randomness is thus used in research to help individuals who find it painful to make a produce an unbiased result. This is a sort of decision. fairness, not to the volunteers but to the drugs Among friends, random selection will be or theories being tested. seen to be fair only if everyone agrees to it. As in all situations, fairness is in the eye of the Random Drug Testing beholder. Often a better method is allocation on the basis of need or appropriateness. In some workplaces, employers require Randomness is likely to be seen as fair only workers to be tested for use of drugs, typically when needs or desires are seen as equally by analyzing their urine. There are various valid, or when no one really cares about the ways this can be implemented, for example final choice. before a person is hired, after there has been an accident, or when someone is suspected of Research using drugs. There is a rationale for each of these ways, either to screen applicants, A scientist wants to test a new pain-relieving determine culpability, or target likely users. drug on humans. The task seems straightfor- It is more comprehensive to test all ward: just pick out some volunteers who are in workers, but this can be quite expensive. pain and see whether they obtain relief. But However, even without universal testing, it is it’s not that easy. The first complication is that possible to have a similar impact with random the volunteers might report reduced pain testing: any worker at any time may be tested simply because they believed that the drug on the basis of chance. Since workers never was helping. A placebo — a pill without the know for sure when they might be tested, they drug — might give just as powerful an effect. are inhibited from using drugs nearly as much The way around this is to give some volun- as with universal testing. Random testing in teers the drug and some the placebo, and see effect operates as a form of universal surveil- whether the ones getting the drug have more lance over the workforce. For the same reason, Random selection in decision making 13

workers are much more opposed to random but this has been resisted by many sporting testing than they are to testing before bodies. employment, after accidents, or on suspicion To say that random drug testing is an (Gilliom 1994: 64). effective way to keep surveillance over a “Random” drug testing implies fairness: no population’s drug use is not necessarily to one is singled out for attention, so to be endorse drug testing, nor indeed to agree that selected for a test implies no guilt. The more drug use is undesirable. In the case of workers, obvious the randomness, the fairer the system for example, it can be argued that drug use is seems to be. If in a workplace the selection of only bad if it creates a hazard or significantly workers to be tested is made by a computer or reduces productivity, and that functional lottery draw, the decision may be seen to be testing, such as using video games to test fair (even if unwelcome). This level of fairness workers’ alertness and responses, is more is not always possible. In some countries, relevant to the stated goals of making the police have the power to stop drivers and ask workplace safer and more productive. Fairness them to have their alcohol levels measured by in many sports might be better achieved by a breath tester even though there is no suspi- providing equal equipment — such as standard cion of alcohol consumption or unsafe driving. racing bicycles — and equal access to drugs This is called “random breath testing.” In (Bakalar and Grinspoon 1984). Our point is practice the tests are anything but random. that randomness can be used to make a system They are likely to be held at locations and fairer within its own terms, without necessar- times of the day when drinking driving is more ily endorsing the system. If competitors are to common — certainly not during the rush hour receive standard bicycles, then they had better for commuters going to work. The testing be allocated randomly. might be more accurately named “somewhat random selection of individual drivers from Lotteries with Bad Prizes targeted times and places.” The label “random” is brief and captures one element of Normally people think it is a good thing to win what is going on. a lottery, but sometimes the “prize” is Whether drug testing actually achieves something that everyone wants to avoid. In the what it promises depends enormously on early 1970s, the U.S. government used a implementation. One group that is subject to lottery, based on birthdays, to decide which drug testing is athletes. Some testing is young men would be drafted into the army; announced before events, allowing individuals many draftees were sent to fight in the war in to drop out of the competition due to spurious Vietnam. The purpose of this system was to be injuries or other excuses. In other cases only fair in the sense that every eligible male had winners are tested, allowing current drug users an equal chance of being picked. (In practice, to avoid testing by hanging back. Although perfect randomness in choice of birthdays was only one or two percent of athletes test not achieved in the 1970 lottery (Fienberg positive for banned drugs, informal opinion is 1971).) However, many people opposed both that 10 times this proportion actually use the war and the draft, and from their point of banned drugs at one stage or another, depend- view the allocation, however fair within its ing on the sport (McGuire 1990: 12). Some own parameters, took place within an unfair drugs cannot be detected by testing and there system. Draft lotteries were also held in the are numerous ways to cheat by going off drugs U.S. during World Wars I and II (Fienberg a suitable time before testing or by using 1971), as well as in a number of European masking drugs so that tests are not effective. countries. Random surprise testing at any time (whether In November 1998, a lottery was held in the during or between events) would be far more Australian state of New South Wales to decide effective in detecting and deterring drug use, which loggers would lose their contracts and hence their jobs. The NSW government had 14 Random selection in politics

increased the area of national park in the high scores on tests in mathematics or English southern part of the state, a major timber do not guarantee that a student will have the region, thus reducing the amount of pulp wood dedication, principles, or human concern to that could be harvested. As a method of make a good doctor or lawyer. Indeed, it might cutting back on logging crews, the forestry even be argued that those who go into these multinational corporation Harris Daishowa professions mainly for status or money are not used a lottery. A front-page story in the the best choices. Sydney Morning Herald (Woodford 1998) In the Netherlands, a different method is emphasized the unfairness to the logging used. Anyone who obtains a high school result contractors who lost in the lottery. considered good enough to undertake a degree However fair it may be, random selection is considered for entry. If there are too many may become tainted when it is used to allocate candidates for a particular course that is in “bads” such as military service or job losses. high demand and expensive to run — such as The taint will be exacerbated when participa- medicine, dentistry, and veterinary science — tion in the lottery is not voluntary. entry is determined by random selection from qualified applicants. For example, suppose University Entry you want to study dentistry, but there are twice as many students who want to study dentistry In most countries, entry to university is as there are places. You enter the dentistry determined largely by merit as determined by course lottery. If you are successful, you get to high school grades and performance on undertake the course. If you are unsuccessful, standardized tests. Scandals occur when you can try again next year. Those who don’t cheating is detected, for example in the U.S. care all that much may decide to study Scholastic Assessment Tests. There may be something else. Those who have their heart set debates about the appropriate tools for on dentistry will keep applying, and most of choosing between applicants but little debate their numbers eventually will come up. about reliance on the merit principle. Debate The Dutch system is a response to the becomes heated when there are deviations problem of selection error, namely that results from “merit,” for example affirmative action in national examinations are not a perfect policies that involve quotas. Merit is also predictor of success at university. However, violated when certain students are given easier students who had high scores in the examina- entry because they are children of graduates, tions are given increased odds of success. live in specified regions, or fit ethnic, gender, Figuratively speaking, good results can be or religious categories. traded in for extra tickets in the entry lottery, Some courses of study are far more popular but every student who reaches a specified than others, often because they are perceived minimum standard has at least one ticket as leading to better jobs. In Australia since the (Riekele Bijleveld, personal communication, 1980s, there has been intense competition to 1993). gain entry to undergraduate courses in University entry by merit and by lottery can medicine, law, and, to a lesser extent, be considered to be based on two different commerce. The result has been that very high conceptions of fairness. Merit-based allocation scores on standardized tests are required to operates on the assumption that the “better” gain entry to these degree courses. candidates — those who are more talented or From one point of view, this is only fair: higher achieving — should have their career those who do better in high school should have choices satisfied before those who are rated the first opportunity to take the university lower. Lottery-based allocation, in contrast, course of their choice. But from another point operates on the assumption that all candidates of view, it is unfair to students who could be who satisfy a specified minimum requirement good doctors or lawyers but who do not score — those who are good enough — have an high enough on standardized tests. After all, equal entitlement to satisfying their career Random selection in decision making 15

choice. Alternatively, random selection can be (This would also provide an incentive to justified as a screening device for entry to improve the poor end of the offerings.) desired faculties when there is no conclusive Goodwin proceeds to make the case for a and fair way to distinguish between the similar distribution of other “goods,” such as applicants. travel opportunities and meals at expensive As well as assumptions about individual restaurants. Perhaps the most eye-opening entitlement, there are important social option is random distribution of job opportu- implications associated with the different nities, which could be redistributed at regular allocation principles. Merit-based allocation intervals. A host of objections spring to mind; leads to a concentration of “better” students in Goodwin deals with them all. Objections come highly desired courses and occupations, from socialists but especially from liberals whereas lottery-based allocation spreads these who support allocation on the basis of merit. students across a wider range of areas. We do not endorse Goodwin’s proposal but Although lotteries have occasionally been find it thought-provoking. used for university admissions in the U.S. However, to argue on the basis of merit or (Wolfle 1970; see also Jump 1988), in some other principle of “deservingness,” such countries where merit-based allocation is well as hard work or loyalty, does not avoid the established, there are few voices advocating role of chance in deciding who gets what. lotteries. Goodwin points out that a single chance event, If the goal is equality, then a complete namely birth, determines much about our solution is equal provision. However, when lives, especially our parents and our natural goods are limited — as in the case of entry to endowments. “Genetic chance” thus influ- certain university courses — then equality is ences abilities and attitudes. Why, Goodwin impossible. If equality of outcomes is asks, should this be seen to be fair? Is it not impossible, then it can be argued that the next fairer to give everyone a chance, at routine best solution is equality of chances — selec- intervals, at different jobs and social rewards? tion by lottery (Oppenheim 1977). The Dutch system thus places a higher premium on RANDOM SELECTION AS A equality of chances than do entry schemes STRATEGY TO MAKE DECISIONS based on test scores, high school grades, family connections, or quotas. Imagine that you are helping organize a nonviolent protest action in a small town. It Justice by Lottery could be about taxes on farms or safety at schools — the topic doesn’t matter. Suppose Barbara Goodwin (1992) in her fascinating there are two obvious places for the action: the book Justice by Lottery makes a case for civic hall and the square. A small group of fostering equality in society by distributing all opponents want to disrupt the protest by major opportunities and rewards by lottery. occupying the location beforehand. They Consider housing. In the short term, there is no suspect you will choose either the civic hall or prospect of achieving equality since there is the square but don’t have enough people to such a huge investment in present housing cover both locations effectively. If the stock, from mansions to tiny units. So instead opponents choose the same location as you do of trying to equalize the overall provision of and prevent the action, they win and you lose. housing stock, Goodwin advocates allocating If the opponents choose the wrong location, equal chances within the existing stock. For they lose and you win. How do you decide? each five-year period, for example, people If the opponents are knowledgeable, they (individuals or families) might be allocated to will look at previous actions and try to work housing by lottery, so that everyone would out a pattern. If they have informers, they will have a chance to live in a nice house in a plush find out about your deliberations. To suburb or to end up in less salubrious quarters. overcome this, one method is random 16 Random selection in politics

selection. At the last possible moment — Some polls are obviously biased. One big perhaps just as your protest group is assem- problem is leading questions. The whole topic bling — you flip a coin to decide whether to of survey and questionnaire design is vitally go to the civic hall or the square. This will important and much studied (Payne 1951; give you a 50 percent chance of winning no Sudman and Bradburn 1982; Sudman et al. matter what the opponents do. Even if the 1996). opponents know your strategy, they cannot do There are other problems, too. A radio better (or worse) than 50% because they station says “Are you for or against the new cannot predict which way the coin will land. tax on land worth over $1,000,000?” and asks This example is artificial, to be sure. Never- listeners to call one of two numbers to register theless, it captures a key feature of many their votes for or against. These sorts of polls decision-making situations in which one’s best suffer from the problem of self selection. Only choice depends on the choice of an opponent, people who happen to hear about the poll and and vice versa. By making one’s own decision are concerned enough about the issue are randomly according to probabilities worked likely to register their votes. Those who “vote” out depending on choices and outcomes, a are unlikely to be representative. Million- specified average return can be guaranteed no dollar land owners are likely to vote repeat- matter what one’s opponent does. This sort of edly whereas others are unlikely to be so strategic use of random selection applies in enthusiastic, especially if there’s a service various areas of economics, international charge for each vote. relations, politics, and war. For example, The ultimate opinion poll involves asking companies need to develop their own invest- everyone’s opinion. This is very expensive ment or sales plans in light of likely moves by and difficult. The closest approximations are competitors. Army commanders need to censuses (which collect personal data rather develop plans of attack and defense in light of than survey opinions) and elections and likely moves by the enemy. referendums with a high turnout. There is a sophisticated branch of mathe- To obtain an approximation of the state of matics dedicated to solving these sorts of opinion throughout the entire population, it is problems, called “game theory.” There is no standard to use random selection to pick out a need to go into details (for readable introduc- sample of the population. Let’s say there are a tions see Davis 1970; Luce and Raiffa 1957; million people. By surveying just a thousand, Rapoport 1960). The main relevant point here a very good approximation of the balance of is that the use of random processes as a opinion can be obtained on most issues, such strategy to make decisions is for the purpose as “Are you for or against the death penalty?” of winning. It has little to do with fairness, or “Do you believe that the earth is visited by except in the tenuous sense that it is “fair” for spaceships from other worlds?” Suppose that the better competitor to win. one out of ten people believes that aliens visit the earth. It could be, just by chance, that all RANDOM SELECTION TO JUDGE thousand people surveyed happen to be OPINIONS believers in aliens, giving a false result of 100% believers. But this is so unlikely that the Opinion polls are an everyday part of politics pollster has more chance of being hit by a and political commentary (Asher 1995). meteor. Statisticians can work out the Pollsters find out political opinions and voting likelihood of deviations from the true percent- intentions, but a larger part of their work is age. It turns out that even for a large popula- commercial, asking whether people recognize tion, quite a small sample can give a fairly the names of car tires, use certain brands of accurate result nearly all the time. cosmetics, or are familiar with certain Consider a population of one million magazines. Nevertheless, we’ll focus on polls people, some of whom support policy A, while about social and political issues. the others support policy B. How many people Random selection in decision making 17

would you need to poll before you obtained an Questions can also be directed to respondents accurate estimate of support for A and B? If as self-interested individuals rather than you ask 100 people, randomly selected, and 50 responsible citizens (Barber 1992). An opinion of them say A, then you can be 95% sure that poll can be fair in the sense of being unbiased, the actual support for A is between 40% and but beyond that the concept of fairness isn’t 60% of the population. If you ask 2500 people really relevant, since a poll is normally not a and 1250 of them say A, then you can be 95% method of reaching decisions. However, sure that the actual support for A is between policy makers may base their decisions on poll 48% and 52%. The result is virtually the same results, in which case the role of randomness whether the population is one million or 100 in ensuring an unbiased result is crucial. million. The concept of a poll can be expanded to Accuracy can be improved by choosing the include providing information, inviting delib- sample carefully. If everyone chosen comes eration, and fostering interaction among those from the same suburb or has a low income, the whose opinions are sought. This can be called result will be biased. So pollsters make sure a “deliberative poll.” For example, citizens that they pick a sample that has the same might be given detailed information about characteristics as the population as a whole: options and encouraged to reflect before the same percentage of males and females, the providing their views. How such methods same geographic spread, the same range of influence the quality of choices made is not education levels, the same range of incomes, easy to determine (Price and Neijens 1998). the same ethnic mix, and so on. This is called We discuss a number of these approaches in a stratified sample. Let’s say a key variable is Chapters 4 to 6. Once deliberation and sex, in that women are likely to answer interaction is introduced into a poll, the differently from men. Then it is vital to stratify process does far more than survey opinions: it the sample by sex. Suppose that women are fosters education and participation. 52% of the population. Then to make up a sample of 100 respondents, 52 would be RANDOM SELECTION TO CHOOSE chosen randomly from the women in the DECISION MAKERS population and 48 from the men. In other words, within strata, cases are selected The Jury randomly. With such stratified samples, statisticians can work out the likelihood of The jury is the most well known example deviations from the true result. today of the use of random selection to choose Actually, this isn’t the most efficient way to decision makers. In a court case, a jury is proceed. A more accurate result can be selected from the citizenry to hear evidence obtained by making some samples proportion- and arguments and reach a judgement. ately larger or smaller than their percentage of Criminal juries are used widely in the English- the population. For example, suppose women speaking countries, especially the U.S., but are have a variety of opinions but men all answer far less common elsewhere. Juries are most the same. Then it would be necessary to ask likely to be used for serious criminal cases, just one man his opinion; the rest of the though they are sometimes used for civil cases polling would be of randomly chosen women. too (Abramson 1994; Enright and Morton The trick then is to combine the two samples 1990; Finkel 1995; Hans and Vidmar 1986; with appropriate weightings, such as 48% and Kalven and Zeisel 1971; Simon 1975; Zerman 52% in this example. 1981). In the case of opinion polling, random Why should jury members be drawn from selection is one of the techniques used to help the population? On the surface, there is a obtain an accurate result. The result can be strong case for sticking with judges. After all, biased in various ways, for example by asking most of them are trained in the law and have leading questions or ignoring certain options. great experience with it. They are familiar 18 Random selection in politics

with the tricks of the legal trade and learn a lot (Enright and Morton 1990: 17-18). In the U.S., about criminals and their stories as well as until recently few jury panels were fully about police and prosecutors. In short, judges representative of the adult population, since have expertise and experience. Why shouldn’t jury source lists were commonly formed solely they do the job rather than a random group of from voter registration lists, and a substantial citizens with no particular expertise and little proportion of citizens are not registered to vote or no courtroom experience? (Piven and Cloward 1988). (This problem Many judges are honest, hard-working, does not arise in the many European countries knowledgeable, and wise. But no matter how that have compulsory voter registration.) good they are, it can be hard to avoid the To overcome this problem and thus help appearance of bias and links to vested inter- eliminate ethnic and other imbalances, jury ests. A person before the court may suspect source lists are now often compiled using that the judge has a “special relationship” with multiple lists, especially driver registration the police, or has been paid off by the records. Other approaches include the use of prosecution. Unfortunately, this is true of different sampling methods, such as cluster some judges. sampling and stratified sampling, and the ad A key reason for juries is for justice to be hoc addition of minority individuals to jury seen to be fair. The jury members are panels. Another bias is caused by the inability independent of the legal system. Even if one or reluctance of many people to serve on or two of them are compromised or biased, juries, due to work or family commitments or there should be enough honest members to just distaste. The “yield” of actual jurors from ensure fair play. jury panels can be as low as 10 percent. The Juries have a long history of being yield can be improved by removing occupa- independent. In the early days of the jury in tional exemptions from jury service, providing Britain, the sovereign sometimes put pressure better pay for jurors, and by a rigorous process on juries to change their verdicts, even to the of pursuing prospective jurors through a point of putting them in prison and withhold- summons, follow-up letters, and allowing jury ing food. The refusal of juries to acquiesce service to be postponed (Domitrovich 1994; was crucial in establishing the jury as an Fukurai et al. 1991; Munsterman and independent body beholden to no one. Munsterman 1986). Juries can be selected randomly from Although subject to limitations, random citizens or in some nonrandom way. However, selection has the advantage that it can’t be any nonrandom process opens the door to bias controlled by anyone and that this is obvious and accusations of bias. For example, in to everyone. Therefore some level of fairness Denmark, committees of politicians choose is achieved and seen to be achieved. jurors from lists of local members of political As well as fairness, there are other ration- parties, thus immediately excluding 95% of ales for the jury. One is that a person charged the population. In addition, the selectors, who with a serious crime should be judged by work down a list of party members, may use peers. Judges, sitting in a position of power their knowledge of individuals, for example and privilege, may get out of touch with whether they are known for having “humanis- “community values.” Juries bring justice back tic” viewpoints (Marcus Schmidt, personal to what “ordinary people” think is right. communication, 17 February 1999). Juries also serve a powerful function Even when random selection is the official politically in defusing anger towards govern- procedure, juries have never been perfectly ment officials. Some court cases are incredibly randomly chosen from the entire local adult charged politically, such as murder trials population. For example, it was only in 1972 concerning horrific crimes, and whatever that Britain eliminated the property qualifica- decision a government or judge would make tion for jurors, thus allowing millions of could unleash potent forces of anger and propertyless citizens to be placed on jury rolls resentment against the government or judge Random selection in decision making 19

directly. A jury helps to defuse these according to their own criteria. In one emotions, since it is composed of many Australian example, the Queensland premier individuals — none of whom can be entirely (equivalent to the governor of a U.S. state) blamed for the jury’s decision — who are was prosecuted for perjury. The premier, Sir “ordinary citizens” who have no special Joh Bjelke-Petersen, was a high-profile leader connection to the defendant or prosecutor. of the National Party. The jury decision was Random selection has always been used in discredited when it was revealed that the conjunction with various forms of screening. foreman of the jury was Luke Shaw who held In many jurisdictions, prosecutors vet potential an official position in the Young Nationals, the jurors: police run checks to eliminate youth branch of the National Party. individuals who are ineligible because of The role of the jury has come under attack criminal records. Of greater concern, in Britain from a number of quarters. Corporations and spy agencies have been involved in vetting on their advocates have argued that juries are the basis of legal political activity (Enright and inappropriate in cases involving complex Morton 1990: 38-52). The prosecution and technical issues, such as environmental cases defense can quiz potential jury members and involving specialist expertise, cases involving rule them out on specified grounds and, to intricate corporate structures and alleged some extent, without giving grounds at all. In corruption, and cases involving medical the U.S. this process of voir dire can be a treatments where experts disagree. Britain and grueling ordeal that eliminates many potential some U.S. states have legislated to allow jurors. smaller juries and majority verdicts, in which If some jurors are biased or have a conflict verdicts can be reached although one or two of interest, the defendant is protected by the jurors disagree, rather than unanimity. These requirement for unanimity, since at least some changes usually have the intent of making jurors may be independent and fair minded, decisions easier to achieve and reducing costs though they also have to be resolute enough to from hung juries. However, they also have the resist strong pressures to acquiesce in the effect of reducing the credibility of the jury’s search for jury consensus. Possibly the most verdict. controversial reform to the jury system in A key accusation is that juries often make Britain in the past century was the introduction the wrong decision, due to ignorance, collec- of majority verdicts — requiring agreement of tive bias, susceptibility to persuasive lawyers 10 of the 12 jurors — in 1967 (Enright and or witnesses, or just laziness. However, it is Morton 1990: 69-76). difficult to back up these charges with empiri- Thus the jury is set up to be fair and be seen cal evidence. The fundamental problem is to to be fair and to provide extra protection to determine what the “correct” verdict actually innocent defendants against conviction. This is is. A number of studies have compared jury in accord with a value judgment often verdicts with the judges’ personal views — expressed as “it’s better to let nine guilty obtained through questionnaires — about what people go free than convict one innocent they believe would have been the correct person.” verdict. The results show that judges and juries In spite of all the safeguards, juries are don’t differ all that much (Kalven and Zeisel hardly free of criticism. Sometimes cases are 1971). It appears that the primary determinant so notorious and polarizing that a jury of the jury’s verdict is the evidence rather than decision, whatever it is, cannot convince or bias, oratory, or laziness. But what if the judge mollify the population. This was seen in some gets it wrong? That certainly happens famous U.S. cases, for example the acquittal sometimes. There is no ultimate way to of police officers who beat Rodney King and determine whether juries are making mistakes, the acquittal of O. J. Simpson on the charge of but the evidence that does exist suggests that murder. Nevertheless, the situation is far they do pretty well. worse when juries are perceived to be flawed 20 Random selection in politics

There are some good reasons why juries fairness in one aspect of justice systems in can be expected to make sound judgments. English-speaking countries, it is embedded in Their inexperience with court cases is a wider adversarial system based on plea balanced by a freshness and openness to the bargaining that is far from fair for many evidence. They are less likely than a judge to accused (Langbein 1992). carry preconceived ideas about a particular defendant or prosecutor. Individual jurors may Classical Athenian Democracy lack specialist expertise, but most juries in industrialized countries include people with In the fifth and fourth centuries BC, Athens considerable knowledge, skills, and judgment. achieved a level of democracy that is still held Being a parent, running a small business, up as a model today. Random selection, working in a corporation or government usually referred to in this case as the lot or office, or being involved in community sortition, was a central feature of classical activities each provide knowledge of the social Athenian democracy (Hansen 1991; Headlam world that can be most useful in assessing 1933; Manin 1997: 8-41; Mulgan 1984; see evidence and forming judgments. In addition, also Aristotle 1984; Jones 1960; Sinclair jurors sometimes have skills in areas involving 1988). science, education, economics, or policy. At the time there were perhaps 30,000 to A crucial part of being on a jury is discuss- 60,000 citizens in Athens. The most important ing the issues with the other jurors. This decisions were made in the assembly, which allows each person to express and test ideas every citizen could attend. The assembly was and arguments, to reject poor ideas and work assisted by several other bodies. The business towards a consensus. To be sure, this process of the assembly was prepared by the Council, often is far from perfect. Even so, the mutual which had 500 members, made up of ten testing of ideas is a remarkably powerful sections of 50 each chosen by lot from one of mechanism for improving judgment. A single the 10 tribes of Athens. Their term of office judge does not have this luxury. For juries, it was one year. Each section of 50 in turn is built in. served as a prytany or committee, the order In summary, a key reason for the existence being determined by lot. The person serving to of juries is the need to be seen to be fair and preside over the prytany, the Council, and the avoid bias and conflict of interest. Juries in assembly was chosen by lot on the day it met. practice reach verdicts that are usually the Lot was also used to choose members of same as judges. What they sometimes lack in very large juries — with hundreds or even specialist expertise they make up for in thousands of members — on courts that breadth of experience and the mutual testing decided private cases but often raised wider of ideas. political issues. Many court cases were Although the U.S. Constitution guarantees essentially used to provide a means for a jury trial in all criminal prosecutions, in holding a vote of confidence in a leader. Thus practice almost all cases are settled through court juries served as de facto policy makers, plea bargaining (admission of guilt to a lesser as indeed do contemporary criminal juries, charge or for a reduced sentence). As a result, though in a far less planned and more prosecutors acquire great power, accused attenuated fashion (Jacobsohn 1984). people are placed under coercion, and criminal Most public officials were chosen by lot. statistics are distorted. Juries are used so Various committees dealt with matters such as seldom that citizen participation in courts is letting out public contracts, collecting revenue greater in those European criminal justice and inspecting markets. A typical committee systems where citizens and judges join in a had 10 members chosen by lot for a term of non-adversarial system of gathering evidence one year. Those who had served were not and impartial prosecution. Thus, although allowed to be reappointed to the same office. random selection is useful in providing For example, Athens had 10 treasurers, 10 Random selection in decision making 21

sellers and 10 receivers (dealing with, for with the lot was there genuine rule by the example, mining contracts), 10 city magis- people. trates, 10 market magistrates, 11 who ran the There were considerable controls on jails, five supervisors of road building, 40 officials selected by lot. Those serving in dealing with private law suits, 10 auditors, 10 office had to be over 30 years old. They were to perform sacrifices and run festivals, and subjected to initial assessment on selection nine archons to deal with processions and have and, at the end of their terms, their records initial charge of law suits — all chosen were scrutinized. This process of scrutiny, annually by lot from citizens over the age of among other things, probably discouraged 30. (About two out of three citizens were over some from volunteering for the ballot, 30.) When 10 officials were selected, usually especially those from the lower classes. The one was taken from each of the 10 tribes. strict system of auditing and the subdivision of The only exceptions to sortition were duties meant that the discretionary power of certain offices where competence was consid- officials was limited. ered absolutely vital. This included many One advantage of the lot for ancient military officers — generals, squadron Athenians was reduction in factionalism and commanders, regimental commanders, and competition for office. This was the reason for cavalry commanders — and some financial selecting equal numbers from each of the 10 officials. These positions were filled by tribes — a type of stratified sample. Conflict election. However, unlike today’s elections, was focused more on the issues, over which there were no political parties or organized there was vigorous debate, than on the struggle campaigning. for office. Athenian democracy was far from perfect. Another important advantage of the lot was Only citizens were involved in the democratic that it helped keep power in the hands of the activities, and women, slaves, metics (resident assembly, reducing the prospect of acquisition foreigners), and children were not citizens. of power by executives, bureaucrats, and Furthermore, it can be argued that participa- wealthy citizens. Because those selected into tion by citizens was not as active nor as executive and administrative jobs had no welcome as presented in later accounts mandate and because their positions could not (Campbell 1989). Nevertheless, classical be renewed, the central role of the assembly Athenian democracy shows that running a was maintained. Randomly selected officials substantial society using highly participative were more likely to act as delegates of the means, especially random selection, is assembly, since they had no independent feasible. mandate or expectation of continuing in office. Today, most people think democracy means Random selection thus provides a way of elections. Representative government is a preventing the gradual expansion of power by system that highlights equality of opportunity, officials, in the context of popular participa- but not equality of outcomes. In classical tion. The ancient Athenians actually sought to Athens, in contrast, political equality was the make holding office unattractive to power goal: an equal sharing of social goods, seekers. including political office. A degree of ineffi- For most positions chosen by lot, Athenian ciency was tolerated in most areas in order to citizens could not be selected more than once. provide every citizen with time in office. An important consequence was rotation in Sortition eliminated the feeling of entitlement office; no one could build up a power base that comes from election or selection by merit. through staying in a particular position. Members of committees were there because of Because there were so many offices for the chance, nothing more. To the Athenians, the size of the population, nearly everyone who lot was democratic and elections were volunteered for office could expect to serve at considered aristocratic or oligarchical. Only one time or another. This meant that nearly every citizen had the experience of participat- 22 Random selection in politics

ing in government. Athenian democracy, 1998). This procedure involving alternating unlike today’s representative governments, turns of sortition and election was designed to was the closest that any society has come to prevent any family group or other clique in the rule by the people. Council from dominating the selection. Similar but less complicated procedures were Other Historical Examples of Sortition used in Italian cities such as Bologna, Parma, and Vincenza, as well as in Barcelona in Ancient Athens is where sortition has been Spain. Lot was used longest in Venice, until used most comprehensively for choosing 1797. public officials. But there are some other • In the 1500s, a lottery-based procedure examples too (Engelstad 1989; Manin 1997: was used occasionally to select members of 42-67). parliament and borough officers in England. • Other ancient Greek city-states probably • In parts of Switzerland from 1640 to used sortition, but only in Athens is this well 1837, mayors were chosen randomly. This was documented. because political office could be used for • Lotteries were used in Italian city-states to financial gain; the argument was that everyone select the six to twelve members of the city should have an equal chance at this. government, who had very short terms of • In the 1900s in the tiny European country office, as brief as two months. This procedure San Merino, a procedure reminiscent of that in was used longest in Florence, for much of the the old Italian city-states was used to select the period between 1328 and 1530. To ensure state’s two governors from the 60-member quality, all volunteers for office in Florence in council (Aubert 1959: 16). the 1300s had to undergo a process of scrutiny twice, once by a group of aristocrats—thus Arguments For and Against Sortition maintaining a degree of elite control—and once by a body of citizens who themselves Political theorists have not devoted much were selected by lot. The scrutiny helped weed attention to sortition, so there is no well out incompetents and unworthy citizens, as developed body of thought about it. Engelstad well as preventing challenges to the (1989), drawing on historical examples and aristocracy. comments by philosophers, has conveniently • Also in Italian city-states, sortition was summarized the arguments for and against the used as part of a complex process also practice. involving nomination and voting. This was The arguments for sortition can be divided most elaborate in Venice where it was used to into (1) promotion of equality, (2) representa- select their chief magistrate, called the doge, tiveness, (3) efficiency, and (4) protection who had to be chosen from among the several against conflict and domination. First, sortition hundred members of the Great Council. To can be justified on the grounds that it promotes begin, 30 members of the Great Council were equality, in several ways. First, it gives selected randomly, then reduced to nine by everyone an equal chance of being chosen, another draw. These nine then nominated and whereas in elections factors such as funding, elected a new group of 40, each of whom had appearance, speaking ability, threats, and to receive at least seven votes. These 40 were promises play a big role. Sortition also gives reduced to 12 by sortition. The 12 chosen then individuals an equal chance for developing nominated and elected another new group of their character and sense of self-worth through 25, who were reduced to nine by sortition. serving in office. Another aspect of equality is These nine nominated and elected 45 new self-respect. Losers in a lottery have no reason names, who were reduced to 11 by sortition. to blame themselves, nor should winners have The 11 nominated and elected 41, again none any reason to believe they are superior by of whom could have been chosen previously. virtue of being chosen. Finally, everyone has This final group of 41 elected the doge (Knag an equal chance of gaining any material Random selection in decision making 23

benefits from holding office, or suffering election campaigns. It provides no justification losses if that is the consequence. for more than one term of office. Therefore, A second type of argument for sortition is the social costs associated with an entrenched that those chosen are far more likely to be a ruling elite, such as corruption and suppres- typical cross section of the population, with sion of social innovation, are eliminated. The the same sort of distribution according to sex, ancient Athenians favored sortition not age, ethnicity, income, occupation, and so because it promoted equality but because it forth. Unlike in elections to the U.S. Congress, “safeguarded the powers of the people, sortition does not lead to over-representation prevented conflict and counteracted corrup- of male middle-aged lawyers. Of course, tion” (Hansen 1991: 84). sortition occasionally could lead to atypical The arguments against sortition can be choices, by the luck of the draw. Just by divided into (1) lack of rationality, (2) reduced chance, twelve jurors could all be women, or obligation, and (3) inefficiency. First, sortition they might be mostly plumbers. This can be appears to involve an abdication of responsi- overcome by stratified sampling, so that each bility and rationality. If reason can be used for designated category, such as females or people human betterment, then, arguably, it should be with low income, is chosen randomly used to select decision makers. Against this, it according to its proportion in the population. may be countered that it is rational to design a Representativeness in sortition means that the system, such as sortition, that undercuts the individuals are chosen randomly but any panel influence of ambition and special interests. with substantial numbers will have the same The details of how sortition is implemented sorts of characteristics as the general can be determined using reason, although the population. actual choices made are left to chance. Actually, some proponents of participation A second type of argument against sortition criticize sortition for not doing enough to is that it can weaken the sense of obligation of ensure that all types of people in society are those selected. Since they have done nothing brought into decision-making roles. If some to deserve selection, randomly chosen office people decline to be involved in a lottery, bearers may feel no obligation to serve the there is no way they can be selected. Further- general interest. This may be true, at least in more, if certain sorts of people, such as those some cases, although Chapters 5 and 6 provide with low self-confidence or heavy family evidence that casts doubt on this view. responsibilities, are unable or unlikely to The third type of argument against participate even given the chance, then being sortition, and undoubtedly the most frequently chosen in a lottery doesn’t change things. heard, is that it is inefficient because the best Measures can be introduced to deal with some people are unlikely to be chosen through a of these problems, such as providing pay for random process. If venal or psychopathic participants plus help in dealing with children, individuals are chosen, the consequences ill health, and other problems. It is important could be horrific. However, elections do not to note that no other selection procedure, eliminate selection of unsavory individuals. In unless there is forced participation, is likely to any case, this problem can be overcome by do better than sortition. Criminal juries are not putting restrictions on who can be included in an ideal cross section of the population, but the lottery, such as minimum age, lack of are far closer than lawyers or court officials. criminal record, or testimonials from a The third type of argument for sortition is specified number of citizens. Alternatively, that it is efficient. Random selection requires individuals selected could be challenged in a no campaigning and little funding. public meeting if enough citizens signed a The fourth type of argument is protection petition, rather like the recall procedure. Of against conflict and domination. Sortition course, every restriction on who is included in reduces the social costs of conflict, such as the a lottery reduces the representativeness of the polarization of views and false promises in resulting choices. Undoubtedly there is a 24 Random selection in politics

trade-off between randomness and guarantees PERCEPTIONS OF FAIRNESS of competence, at least in the short term. A related criticism is that sortition doesn’t As noted earlier, Goodwin (1992) argues that allow building on the knowledge and wisdom using a lottery is a fair way of allocating all of experienced decision makers. Because sorts of things, such as housing and jobs, that random selection provides no rationale for most people would never think of distributing remaining in office, there is a considerable randomly. Just how fair is random allocation turnover of officials and hence a loss of of goods seen to be? There isn’t a lot of continuity. This problem can be ameliorated investigation of this. Wortman and Rabinowitz by gradually phasing in newly chosen (1979) carried out an ingenious experiment individuals, so that there is not a sudden with hundreds of undergraduate psychology replacement of an entire group. students, comparing their perceptions of four Obviously there is a tension between, on criteria for allocation of a scarce good: merit; the one hand, giving opportunities for a need; first come, first served; and random diverse range of people to participate in assignment. The students were applying to decision making and, on the other, maintaining participate in an innovative educational continuity and quality of decision making by program that they judged to be extremely sticking with the most experienced and attractive. Different groups of students were knowledgeable people. Curiously, we have told different stories about which criterion had heard some individuals argue against sortition been used to make the selection, and whether on both grounds: they are worried both that they individually had been successful. All random selection won’t create a sufficiently were asked which selection criterion was participatory system and that it will lead to a fairest. Random assignment was considered loss of continuity and wisdom. Ultimately, the the most fair. Furthermore, random assign- solution to this pair of problems is to develop ment was perceived as fair regardless of a breadth and depth of knowledge and whether the students thought they personally experience within the population sufficient to had benefited from it. In contrast, students ensure an adequate quality of decisions who were led to believe they had been whoever is chosen. But that cannot be successful due to their merit were more likely achieved in a day whatever system is used. to support merit-based assignment, and In this outline of arguments for and against similarly for the principles of need and first sortition, we have emphasized the arguments come, first served. This suggests that random for and given counters to the arguments assignment is both seen to be fair and best against. Sortition is a neglected option, but not overcomes the self-interest attached to other because the case for it is weak. Rather, we criteria. believe it has been neglected primarily However, this finding does not necessarily because it threatens interests that benefit from apply elsewhere. For those favoring random other selection methods — whether election, allocation, it is especially important to investi- appointment, heredity, or brute force. Random gate situations where it is seen to be unfair and selection is hard for any individual or group to to learn why this might be the case. One study control, hence those with power typically by two economics professors (Frey and ignore or reject it. Arguments such as ineffi- Pommerehne 1993) found that random alloca- ciency may be presented to oppose sortition, tion was considered quite unfair. Hundreds of but we believe these are mainly rationaliza- households in Zurich and West Berlin tions. One good way to see if an alternative is answered a questionnaire about hypothetical any good is to try it out, as discussed in situations. A question for one situation was as Chapters 5 and 6. But few governments have follows (Frey and Pommerehne 1993: 298, ever shown much interest in giving sortition a 301, combining two questions and changing fair trial. the cost to dollars):

Random selection in decision making 25

At a sight-seeing point reachable only on need or merit fairly. Erez concludes that foot a well has been tapped. The bottled people need to have explained the benefits of water is sold to thirsty hikers. The price is random selection in terms of overcoming bias. $1 per bottle. Daily production, and thus Prisoners with some higher education were the stock, is 100 bottles. On a particularly most likely to prefer randomness. Finally, the hot day 200 hikers want to buy a bottle. concept of randomness or chance may be off- Please indicate how fair you evaluate the putting for prisoners, who typically see their following means to distribute the water plight as due to “bad luck.” Reconceptualizing among the hikers to be: random selection as “equal chances” — (a) A price increase to $2 per bottle? emphasizing the equality of opportunity rather (b) Selling the water at $1 per bottle than the unpredictability of the outcome — according to the principle of “first come, may make random selection more attractive. first served”? Another possibility is to avoid the word (c) Selling the water at $1 per bottle “random” altogether, for example referring to following a random procedure (e.g. to give “statistical selection.” to all persons whose surname starts with A Whatever the explanation for these results, through to P)? it is important to note that random selection is (d) The local authorities buy the water for not automatically seen as fair. Perceptions of $1 per bottle and distribute it according to fairness depend on the situation and on how their own judgement? randomness is applied. But another factor is that people need to believe that a lottery is a Answer (b) is seen as fair by 76% of respon- fair system, and that may depend on familiar- dents, (d) by 43%, (a) by 27% and (c) by only ity, discussions of random selection versus 14%. The authors are concerned that the price alternatives, and overcoming vested interests allocation, which economists consider the opposed to randomness. fairest allocation method, was seen as so unfair. They note that random allocation, CONCLUSION answer (c), may have received a low rating for fairness because it is not well known and not In making decisions, there are many factors to considered suitable for serious matters. consider, including fairness, participation, Another problem is that the hikers don’t have cost, speed, and good results. Random selec- the same needs: some will be more thirsty than tion is an especially useful tool when a fair others. Finally, the procedure of selling water process is a prime consideration. This applies to those with surnames A through P may not when random selection is used to make even seem random, but arbitrary. decisions directly, as in the case of lotteries, to Erez (1985) asked hundreds of US prison- survey public opinion, or to select decision ers how best to select inmates who would be makers. On the other hand, fairness is not able to take a beneficial course. Four methods relevant when random selection is used to of selection were given: need; merit; first work out the best strategy. come, first served; and random selection. The If everyone can be involved in making a strong first preference was the criterion of decision that affects them, then this is seen as need. The others were far behind, with random both fair and participatory. This can work selection the least favored method. Erez notes when the group is small, but for large groups, that prisoners may prefer the criterion of need the cost and time involved make full partici- since they see themselves as having great pation impossible. That means that only some need. Interestingly, prisoners distrust each people can be involved in any given decision. other’s motives and distrust staff, so they are The question is, who? If the decision makers suspicious of other inmates who claim great are appointed, then a bias is possible. If they need or who volunteer quickly (“first come, are chosen randomly, as in the case of the jury, first served”) and do not trust staff to assess then bias is limited. 3 Direct democracy

“Democracy” is a concept that means many While direct participation in decision things to many people (Lummis 1996: 14-19). making is the key characteristic of direct It is generally seen as a good thing, and democracy, historically it has several other therefore people want to describe their associations. In ancient Athens, as discussed in preferred system as “democratic.” Political Chapter 2, decision-making forums were: scientists say that the concept of democracy is • face-to-face meetings; “essentially contested” (Connolly 1983), • deliberative (namely, issues were dis- which means that the rules for using the cussed and debated by participants); concept are open to varying interpretations and • concerned with issues affecting partici- there is no definitive way of resolving pants. disagreements over usage. In practice, the Another example is town meetings in New meaning of “democracy” cannot be agreed England (northeast United States). In these upon because contesting groups have interests meetings, any citizen can attend, make contri- in different meanings. butions to the discussion and vote on motions. To help pin down a meaning, it is conven- In this ideal form, these meetings are reminis- tional to preface the word democracy with a cent of classical Athenian democracy. Similar descriptive word. For example, “electoral to this is self-management in neighborhoods democracy” and “representative democracy” within cities such as Washington, DC (Kotler refer to political systems in which people vote 1969; Morris and Hess 1975) and self- for representatives who make decisions. government in villages in India and Sri Lanka, “People’s democracy” is a bit trickier. It has called sarvodaya (Kantowsky 1980). In towns, been used by state socialist governments to neighborhoods, or villages, formal meetings describe themselves, such as the German are only a part of what is needed for direct Democratic Republic (the former East democracy. Also vital are discussions in the Germany) and the Democratic People’s community, appropriate education, measures Republic of Korea (North Korea). These to ensure that everyone can participate, and communist regimes consider themselves attitudes such as tolerance and community democratic because the ruling communist spirit. party is supposed to be the embodiment of the Direct democracy is also possible in indus- people’s interests and aspirations. Obviously, try. There are many examples in which it pays not to take every label of democracy at workers have taken full control over their face value. workplace and run the entire operation, with “Direct democracy,” also called “participa- decisions made in general meetings. In some tory democracy,” is a mode of self-rule in instances, such as the Scott Bader company in which people are directly involved in decision Britain, this arises from a benevolent owner making. This form of democracy is “direct” who fosters participation by workers and because people are directly involved; it is eventually turns control over to them. In other participatory because people themselves cases, workers take control in order to prevent participate in making decisions. In contrast, collapse of the company and loss of jobs, as in representative government is “indirect” the “work-in” in two of the four yards of democracy since people choose representa- Upper Clyde Shipbuilders, a government tives who make decisions and is nonpartici- enterprise, in 1971-1972. The workers took patory since people do not directly participate over allocating jobs, arranging insurance, in making decisions. supplying orders, and providing pay for over a Direct democracy 27

year until the British government made the resulted from the collapse of the established yards viable through grants (Coates 1981). government and survived until old or new This is an instance of what can be called forces of domination crushed them. For “industrial democracy.” However, the con- example, the Paris Commune of 1871 tested meanings of democracy come to the involved running the entire city for many fore in this area. “Industrial democracy” is months. Meanwhile, the French military often used to refer to systems in which commanders — defeated in the war with workers have some limited say in how work is Prussia — retreated to the countryside, disci- done or where a few workers’ representatives plined the troops, and then attacked the sit on the board of management. Industrial Commune ruthlessly, killing thousands of democracy does not have to be direct people. The soviets (workers’ councils) in the democracy. When workers run everything at a early stages of the Russian Revolution were workplace, this is more commonly called eventually crushed by the Bolsheviks. In “workers’ control” or “workers’ self-manage- general, it can be said that direct democracy ment.” In any case, the point here is that direct on a significant scale is an enormous threat to democracy can occur in workplaces (Dahl the established order. Governments make 1985; Hunnius et al. 1973; Roberts 1973; every effort to smash such outbreaks of self- Thornley 1981). management. The most long-lasting experi- The most dramatic examples of direct ence of revolutionary direct democracy was in democracy are during revolutionary situations, Spain from 1936-1939, which was the when workers and citizens take control over culmination of decades of organizing. The decision making in defiance of existing anarchist collectives were opposed by both the governments. These instances are often called Spanish fascists (supported by Hitler) and the “self-management,” and the self-governing communists (supported by Stalin) and eventu- bodies may be called communes, soviets, ally crushed. councils, or committees. Some examples are At the other end of the spectrum is consen- (Guérin 1970; Root & Branch 1975): sus, a form of direct democracy in small • communes in Paris in 1792-1793 during groups. Using consensus procedures, there is the French Revolution; no voting. Instead, options are formulated and • the Paris Commune of 1871, in the after- discussed until everyone agrees — or nearly math of the Franco-Prussian war; everyone, depending on procedures. Since a • soviets in the early stages of the Russian single person can block consensus, no one’s Revolution in 1917-1918 (Anweiler 1974); views can be dismissed simply by outvoting • workers’ councils and soldiers’ councils them. Consensus has the advantage of in Germany in 1918-1919, in the aftermath of harnessing the commitment of group World War I; members, since when an agreement is reached, • collectives in Spain during the revolution there is a maximal degree of support, unlike and civil war of 1936-1939 (Dolgoff 1974; voting where the minority who are outvoted Richards 1983); may continue to oppose or even undermine the • factory committees in Hungary during the decision reached (Avery et al. 1981; Coover et 1956 uprising against Soviet rule; al. 1981; Gastil 1993; Kaner 1996). • action committees in France during the When using formal processes of consensus, uprisings in May-June 1968; there are detailed procedures and much • self-managing groups in Chile under the practical experience to ensure that everything Allende government in the early 1970s (Raptis works as well as possible. Formal consensus 1974). procedures have been used extensively in the These instances of direct democracy were nonviolent action movement, the feminist far-reaching, in that they involved large movement, and the environmental movement, sections of the population running major among others. enterprises or entire societies. Many of them 28 Random selection in politics

Consensus works most readily in small OBJECTIONS groups, such as affinity groups (close-knit mutual support groups) of say 5 to 15 people. Direct democracy is not perfect. New England It can also be used for larger groups. One town meetings, workers’ control, and consen- system is for each affinity group to select one sus each have limitations. In addition, direct delegate or “spoke” to a group of spokes, who democracy is threatening to those exercising try to reach consensus. But every affinity or seeking to exercise power, such as politi- group must agree to the decision, so there can cians, top bureaucrats, managers, and authori- be many stages of discussion within affinity tarian individuals. Various objections have groups and by the group of spokes. By this been raised against direct democracy, often means, consensus procedures can be used with with the purpose of justifying unequal as many as a thousand people. participation and power. It’s not always easy Consensus is also widely used on an infor- to separate the genuine limitations from the mal basis, especially in small groups. A group objections that serve to bolster inequality in of friends, in deciding where to go out to eat, power. may seek to find a restaurant to which no one • Objection 1: people are not capable of objects. The same applies in many situations running their own lives. The argument here is in which voting is the official procedure. A that experts and managers know better what is small committee may discuss a topic and only good for the population. Direct democracy, it take a vote as a formality when it is obvious is said, threatens to put those who are ignorant that everyone agrees. This is in tacit recogni- and prejudiced in charge, which would be tion that a vote can be destructive if a disastrous. minority, even a single individual, is strongly This objection is often made by experts and opposed to the resolution. The harmony of the managers. It is a self-fulfilling prophecy, since group, and the good will of every individual, is when people are not given the opportunity to important for the long term success of the participate in decisions affecting their lives, group. It is simply not worth alienating anyone they never develop the capability to do so and — especially a valued member of the group — may even reject the opportunity when it is over a routine issue. In many cases there is offered. informal canvassing of views before the People can become quite expert about meeting to sort out a resolution that will things that affect their own lives, as shown by satisfy everyone. When it comes time to vote the remarkable knowledge acquired by some (“All in favor — all opposed — carried,” with people in self-help groups. With the Internet, it scarcely time for a hand to be raised), hardly is becoming easier to gain information about anyone takes notice since the result is fully all sorts of specific topics. In contrast, politi- accepted. But if a respected member expresses cians know very little about most of the issues vehement opposition, then everyone will come on which they vote. to the alert, because things are no longer • Objection 2: direct democracy takes too routine. Voting is the official method of much time. The claim here is that a single boss decision making, but informal consensus is the or elite group can make the quick, hard underlying or shadow process. decisions, whereas consensus or a workers’ Consensus is participatory and is normally group might take forever, and that people face-to-face, deliberative, and concerned with simply don’t have or want to spend the time to issues affecting the participants. Thus it is make direct democracy work. definitely a method of direct democracy. In its It’s true that direct democracy often takes a informal form, it is undoubtedly the widely lot of time — many participants have used method of direct democracy in the world. complained of this. On the other hand, although decisions by managers may seem quick and easy, sometimes they lead to long- term problems that require lots of work and Direct democracy 29

agony later. Sometimes a slower process is people is limited by constitutional and more efficient in the long term. procedural mechanisms to ensure that basic If participants are volunteers, the time, liberties are maintained. Direct democracy effort, and cost involved can be counterbal- could be a threat to freedom of speech or anced by the sense of doing something freedom from imprisonment without trial, for worthwhile for the community. In society example. People might vote to legalize or today, the number of volunteers who are quite outlaw capital punishment, abortion, drugs, willing to help those in need is testimony to etc. — take your pick! the willingness, even eagerness, of many It is intriguing that many on both the left people to make a social contribution. The time and the right are fearful of popular participa- taken in democratic processes is far less of an tion, thinking that “the people” are too obstacle when participants know that their conservative or radical and will misuse their involvement helps toward accomplishing power. In any case, direct democracy is quite something of value. compatible with a system that protects certain • Objection 3: today’s society is too basic freedoms or rights, for example by complex for direct democracy. Noting that requiring large majorities to overturn them, there are many difficult issues, from home just as in representative systems. detention schemes to ozone depletion, the • Objection 6: persuasive individuals may argument is that the ordinary person cannot have an undue influence in direct democracy. hope to participate sensibly in decisions about A single powerful personality sometimes can them — hence specialists are needed. sway a group using consensus or a meeting of This objection is often made by experts workers. The voices of those who are less who have a vested interest in maintaining their confident and articulate may be better heard own special role. Specialists may know a lot through anonymous voting than in face-to-face of technical details, but lots of people can decision making. grasp the basic issues about which decisions Persuasive individuals are a danger in any need to be made. You don’t have to know how system, but much more of a danger when to fly a plane in order to comment on transport power is unequal. A president or prime policy and you don’t have to know how to do minister has vastly more power than the most brain surgery in order to comment on health charismatic figure in a consensus decision- policy. After all, far more politicians are making group. lawyers than engineers or criminologists. • Objection 7: people will only look out • Objection 4: direct democracy won’t work after their own interests. If people make with large groups. Consensus or a town decisions affecting their lives, they may only meeting might work for hundreds or even look out after themselves, for example by thousands of people, but are claimed to be reducing taxes or blocking local developments impractical for populations of millions. This is (the not-in-my-back-yard or NIMBY an argument for having representatives. syndrome), and not consider impacts on other This objection is fine as far as it goes, but it communities, the environment, or future assumes that the scale of political units is generations. fixed. One alternative is to use direct Actually, politicians are even more suscep- democracy on a small scale and to coordinate tible to this problem, for example often the small units by some form of confederation favoring local developments that have a wider or network. Some countries have hundreds of harmful effect. At least with direct democracy, millions of people, but many others have many of the consequences are bound to fall on fewer than a million, and seem just as able to the decision-making community. Unlike provide citizens with security and prosperity. politicians, the population as a whole can • Objection 5: basic freedoms need to be seldom escape the fallout from bad decisions. protected from the whim of the majority. With • Objection 8: no one can be held responsi- representative government, the power of the ble for bad decisions. Since there is no one 30 Random selection in politics

officially in charge, there is no one who can be conditions in which random processes have held liable for the consequences of decisions the most to offer. made. Certainly the decision makers — everyone — can’t be replaced by a different CONSENSUS set. Actually, since everyone is responsible for Getting everyone’s agreement can be difficult decisions in direct democracy, and the but when it is achieved it can harness community wears the consequences, there is a incredible strength. That is the advantage of strong pressure to reach decisions that serve consensus procedures. However, consensus the community’s interests. In parliaments, has some limitations. One is that in a face-to- often the major parties agree on policies, and face meeting with relatively few people, a there is no way to change the policy through powerful personality may influence voting. Furthermore, the number of politicians proceedings greatly, even though no one has and bureaucrats who pay any significant formal power. A related problem is that some penalty for bad decisions is very small indeed. individuals may find it hard to express their • Objection 9: direct democracy may cause dissent. In voting, with secret ballot, it’s lack of continuity. In ancient Athens, the possible to express a preference that would be assembly sometimes overruled decisions made risky to present openly. In consensus, there is at a previous meeting. Direct democracy could no secrecy. If you object to a proposal, you be unstable and chaotic. need to express your objection. If nearly However, lack of continuity can also be a everyone else agrees with the proposal, it is problem in representative government. There often easier to go along with their views rather is no evidence that it is a greater difficulty in than face the peer pressure to maintain direct democracy. harmony. The result is that powerful person- • Objection 10: it is too expensive for alities can shape the consensus. They have the everyone to participate. Referendums, for confidence to object to proposals they don’t example, are not cheap. However, electronic like, so the group pursues other ideas. The less voting can overcome cost obstacles. Random confident members sometimes don’t object selection is another way of reducing costs, as even when they disagree, so the decisions tend we describe in this book. These objections to go the way the confident members prefer. and others are all worth considering closely. Mansbridge (1980) gives an acute analysis and However, there is a danger in focusing on comparison of both direct democracy in small objections, if it provides a pretext for not groups (which she calls “unitary democracy”) trying out direct democracy. After all, repre- and representative government (which she sentative government has been given ample calls “adversary democracy”). There are also opportunity for a long time, in spite of many less sympathetic criticisms of consensus objections that can be made about it, as (Bookchin 1994; Landry et al. 1985; Ryan described in Chapter 1. 1985). However, our main concern here is not the Another problem in some consensus groups dispute between proponents of direct and is a rigid adherence to formal consensus indirect democracy, but rather the potential procedures. Formal procedures can be helpful, role of random selection in improving but only if they are in tune with the needs of democracy of whatever sort and in leading to the participants. However, this objection is fairer and better decisions. less a criticism of consensus as a method than We now turn to a closer look at several to the way it is run. types of direct democracy — consensus, Because consensus is most commonly used voting in meetings, networks, and referendums in small, close-knit groups, random selection — pointing out some of their strengths and is not likely to seem appropriate. Often groups limitations and highlighting the potential role give each member a turn at being facilitator of randomness. This will help to highlight the and invite comments from everyone present Direct democracy 31

about the matter being discussed. Many important to win over those who may not be advocates of consensus are extremely sensitive expressing an opinion openly. to inequalities of knowledge, experience, and On the other hand, voting disenfranchises confidence, not to mention sex, age, ethnicity, the minority. It ratifies a conclusion in a way and class. that makes it harder to bring up alternatives. It When using consensus in larger groups, can encourage an appeal to emotion or vested random selection might be used to equalize interests to win over wavering voters. Since all participation. For example, when affinity that is needed is a majority, there is little groups choose a spoke, this could be done incentive to pursue the discussion after enough randomly. In this way, the spokes would not people are won over. necessarily be the most articulate or experi- Voting in meetings assumes a face-to-face enced members. This would help to equalize deliberative process, even if only a few people skills and experience within the larger group, have or take an opportunity to speak. Once though at the expense of not fully using the voting is separated from the meeting context, most experienced people at the current the dynamics change considerably. Voting for occasion. Since the “current occasion” is often representatives in an anonymous polling a crucially important issue, such as running a booth, in which there is no face-to-face delib- blockade, there may be resistance to random eration, is quite different from voting in a selection. meeting. Can random selection improve the process VOTING IN MEETINGS of voting in meetings? One idea is that the chair of the meeting be chosen at random The most common model in direct democracy (perhaps from volunteers) at the beginning of is voting in meetings, including in classical the meeting, as done in ancient Athens. In this Athens, New England town meetings, way, it is harder for an in-group to manipulate workers’ control, and revolutionary self- the agenda and meeting procedure. Another management. The usual picture is that the possibility is that people invited to speak be meeting is chaired by some experienced chosen randomly from those offering. For person, participants are invited to speak, example, if ten people offer to speak for the motions are put, and votes are held. motion, two or three could be chosen at If there is a strong commonality of interest, random. This would reduce the influence of such as workers who are united against bosses, pushy, vocal, confident individuals and allow then disagreements are usually not about others a better chance, while being seen to be fundamentals but about matters such as tactics. fair rather than playing favorites. However, even here there can be fierce Sometimes the meeting is manipulated, for disagreement. example by putting a crucial item at the From the point of view of participation and bottom of a long agenda so that it is only dealt fairness, one of the risks in meetings is that a with after many participants have left (but clique will control the agenda and marginalize loyal comrades remain!). Perhaps the agenda dissent. If the vote is brought on too soon, items could be dealt with in a random order. participants won’t be fully aware of the However, there are probably better ways to arguments. If only some of those requesting to deal with this, such as specifying a certain speak are given an opportunity, then the minimum attendance, a fixed time for ending discussion can be unbalanced. the meeting, or voting on which items should Voting has value in that it registers an go first on the agenda. expression of the majority (or some higher Voting in meetings is possible with target, if say a two-thirds vote is needed for anywhere from a handful to thousands of some action). A system of one vote for each people. However, the larger the size, the more person ensures that no individual or small limited the input that the average participant group can impose its will. Voting makes it can make. In other words, interaction is 32 Random selection in politics

reduced. There are various ways to overcome Governments generally seek to maintain their this. One is to break into small groups for part power, and it is harder for governments to of the time to discuss the issues — analogous control the electorate than to control parlia- to affinity groups in large consensus systems ment. The main reason that Switzerland and — and then to return to the full meeting for a Australia have had so many referendums is vote. If the small groups are chosen randomly, because they are required in order to change there is less risk that the process will be seen their constitutions. Many governments — to be manipulated. especially dictatorships — hold referendums As the size of meetings increases, the to provide legitimacy for themselves (Butler participatory aspects decline. It’s possible to and Ranney 1994). Sometimes governments imagine a football stadium filled with 50,000 exclude desirable alternatives and prejudice participants discussing health or transport the outcome by the way the proposition is policy. This may be direct democracy but the expressed. For example, in 1982 the face-to-face and deliberative aspects are Tasmanian government held a referendum attenuated, since most of those attending can giving several options for dam projects, but only listen. For such large numbers, direct omitted the one option backed by a popular democracy requires other mechanisms. environmental campaign: no dams. Random selection can play a role as described One way to overcome government manipu- in Chapters 5, 6 and 7. lation of referendum questions is through the citizens’ initiative. A group of citizens decides INITIATIVE AND REFERENDUM on the proposition and obtains a specified number of signatures endorsing it, after which When citizens vote on a proposition in a the proposition is put to the people as a referendum, this is direct democracy in the referendum. This puts the agenda in the hands sense that they are directly involved in making of citizens rather than officials. The initiative a decision. Compared to decision making by is available in Switzerland and many U.S. legislatures, endorsement by the people in states, where it is regularly used. The initiative referendums gives the results greater legiti- is of vital importance in taking the decision to macy. By bringing issues to the people, hold referendums out of the hands of referendums often generate widespread debate governments, thus reducing the potential for and overcome alienation and apathy. Also, government resistance and manipulation. they are more likely to serve the public interest An important limitation of the referendum compared to decision making by representa- is that deliberation is not built into the process. tives who are more susceptible to pressure However, if a referendum is preceded by from special interests, including interests of debate in the media, formal meetings, and lots the representatives themselves. However, there of informal discussion, then something akin to are a number of obstacles to making referen- the deliberation of a meeting may be achieved. dums a powerful tool of direct democracy. A proportion of the population, perhaps five or In most countries, referendums are only 10 percent, will become well acquainted with held at the discretion of governments. Many the issues — a far higher proportion than for governments have held at least one referen- most issues decided by legislatures. But many dum, but few have held lots of them. At a of those voting may take only a passing national level, there have been hundreds of interest and not have been involved in any referendums in Switzerland and more than 40 interactions about the issue. in Australia. Of all other countries only France When governments run a referendum, there and Denmark have had more than a dozen. is no process of deliberation over the referen- The fact is that most governments do not like dum question itself. This is a crucial holding referendums, and usually do so only weakness. With the initiative, citizens decide when they think they can engineer the result or the way the proposition is framed, but in when they want to offload responsibility. practice only a relatively few actually have Direct democracy 33

input into the wording. This again is a scales of support (“strongly agree,” “agree,” shortcoming from the point of view of “disagree,” “strongly disagree”) and ranking deliberation. of multiple alternatives which has been In representative systems, special interest referred to as multi-option preferendums groups have harnessed the power of opinion (Emerson 1998). polling and direct marketing to pursue their ends. A referendum may seem like an NETWORKS expression of popular opinion, but opinion can be swayed by marketing appeals. This We have talked a lot about face-to-face includes advertisements and “news” resulting meetings in decision making using both from media releases and staged events — all consensus and voting. In ancient Athens, of of which are designed to appeal to citizens on course, interaction was face-to-face or not at the basis of market research — and sometimes all. Today, though, communication technology fake groups to give the illusion of grassroots allows remote interaction, using telephone, support. It is precisely because there is no television, radio, fax, and electronic mail. formal process of interaction and deliberation Meetings can be held that are not physically that referendums are open to manipulation in face-to-face but are otherwise interactive. An this way. However, in spite of the potential for example is a telephone conference, with manipulation, the power of money seems to be several people on the same line. This allows the determinant of the result in only a few participation from people in different loca- cases (Cronin 1989; Schmidt 1989). tions, even different parts of the world. It also Randomness could help to overcome some allows participation from people who might be of this manipulation. Many propositions are unable to physically attend a meeting due to couched as a yes-no question. This makes it expense or disability. Talking on the telephone easy to build a campaign around something reduces some modes of communication, such people can remember easily, such as “Keep as facial expressions and body language, taxes low, vote no on 17.” This means that which carry a large amount of the information people don’t need to know the arguments, just in a face-to-face interaction. Nevertheless, the the answer, yes or no. Imagine, then, that telephone retains vocal expressions. Another referendum ballots are produced in two forms, possibility is videoconferencing, which can one with the question framed so that “yes” replicate much of the experience of “being supports a proposal and the other so that “no” there.” In the future, systems of virtual reality supports it. For example, half the ballots might may make it possible to attend meetings say “I support fluoridation of the town water remotely in ever more realistic fashion. supply. Vote yes or no” and the other half Yet another possibility is an electronic chat would say “I oppose fluoridation of the town group, in which participants type their water supply. Vote yes or no.” Voters would messages and see everyone else’s messages receive one ballot or the other by chance. This pop up on their own screens in chronological might seem only a slight improvement but it sequence. This reduces nonverbal communi- would ensure that people would have to know cation to a minimum, which is a disadvantage what they were voting for or against. Most when trying to get across your meaning referendum questions are not expressed quite (especially subtle concerns). On the other so simply, making the need for voters to hand, email tends to equalize power by understand the question even more important. reducing cues about a person’s formal status An elaboration would be to have the proposi- (Sproull and Kiesler 1991). You may not tion expressed in various ways so that no know whether someone else is a senior official simple how-to-vote slogan or card would be or a high school student. Only what’s written enough to get the “right” answer from the is salient. However, when there are many unthinking voter. It is also worth considering active participants in a chat space, some — options in addition to “yes” and “no,” such as such as slower typists and reflective thinkers 34 Random selection in politics

— may be left behind and frustrated, not need to make the decisions themselves, for unlike the dynamics of many face-to-face example in the manner of initiative and group discussions. referendum. Thinking about the possibilities for decision There have been a number of speculations, making in networks suggests that face-to-face innovative experiments, and criticisms of the interaction is not an essential characteristic of potential for computer networking to improve direct democracy. In its place we can put citizen participation and bring about cyberde- “interactivity” (more accurate would be mocracy (Abramson et al. 1988; Arterton “intercommunicativity,” not a nice word!), 1987; Grossman 1995; Slaton 1992). These namely the potential for everyone to be have many strengths but also some familiar involved in contributing to the discussion as limitations. The impact of information well as being a recipient of other people’s technology depends most of all on the aim of contributions. Electronic networks can serve those who design the systems. It can be used as a new type of public sphere, replacing or for minor reforms in existing representative supplementing traditional civic squares and systems, such as politicians setting up web coffee houses, in which information is shared, sites or pressure groups using email to issues debated, and democracy practiced influence representatives. Direct cyberdemoc- (Tsagarousianou et al. 1998). racy by voting on issues — in effect, a series Decision making in networks can operate of instant electronic referendums — has the using consensus or voting. Random selection advantage of making participation easy. It has can play much the same role as in these the potential disadvantages of limited systems, to select facilitators, speakers and deliberation, limited participation in designing delegates. questions to be voted on, and limited popular control over the format for information ELECTRONIC DEMOCRACY sessions or debates preceding votes. Random selection could be used in various ways to If everyone can be hooked up with a computer improve the process, just as in the case of and modem, then it’s possible to have an conventional initiative and referendum. It is electronic town meeting and a computer vote. important to debate various methods for For example, hearings about an issue could be implementing cyberdemocracy, but the real broadcast on television to a neighborhood, test is to see how they work in practice. city, or entire country. Members of the “virtual audience” (that is, those linked by computer) CONCLUSION could contribute to the discussion by phone or computer video conferencing. The ones Direct democracy has a long and often chosen to contribute could be chosen in exciting history. Because it is a radical various ways, including by lot. The process challenge to rule by elites, it has been opposed could be supplemented by simultaneous chat by powerful interests in both revolutionary and group discussions. everyday situations. Electronic links can be used for various Direct participation is the key to this alter- purposes, including information sharing, native, which is why it is called “direct consultation, debate, and decision making. For democracy” or “participatory democracy.” example, remote viewers and commenters Normally, everyone affected can participate. might be used to provide citizen input on an Therefore, random selection is seldom used: issue before a vote by parliamentarians, or the approach is seen to be fair as it stands. electronic voting might be used to elect Even so, random processes can help to representatives. However desirable they may improve methods of direct democracy, for be, such cases are not direct democracy. To example in choosing facilitators or speakers at have electronic direct democracy, or in other meetings. words direct cyberdemocracy, participants Direct democracy 35

As the size of the group becomes larger, full participation by everyone becomes more and more difficult. One solution is to turn to representation through elections, thus rejecting direct democracy. Another solution is to retain the key characteristics of direct democracy by breaking the large group into smaller groups. Once this process begins, random selection can play a useful role in either selecting the smaller groups or in selecting the smaller groups’ delegates to the full group. This suggests that random selection has a special role to play maintaining fairness in decision making in large groups. The larger the group, the more the danger of insiders and vested interests controlling decisions. Random processes can be used to prevent this, produc- ing a politics that is fair and seen to be fair. 4 Citizen participation without random selection

“[P]eople are almost desperate for this objective facts whereas society is more a opportunity for small-scale discussion. realm of value judgments and power. How- It’s just that it’s dropped out. We don’t ever, in practice, scientists have been heavily have it really anywhere in the culture, involved in technological developments such the church doesn’t do it very much, as nuclear weapons or genetic engineering that service clubs don’t do it very much … have major social implications. that’s the core challenge of the current moment” (Theobald 1998). One idea for dealing with contentious social issues involving science is the “science court,” Direct democracy, as outlined in the previous which would deal with scientific facts sepa- chapter, occurs when people completely run rately from political and value judgments. As the show, whether in a small group, a proposed in the 1970s onwards, it would workplace, a neighborhood, or an entire involve three broad stages: initiation, organi- society. Random selection can be used in a zation, and procedure (Cole 1986: 250). The number of small ways to increase fairness in first stage was to establish administration to some forms of direct democracy such as arrange funding, find a suitable issue, and consensus and voting in meetings. create cooperative relationships with agencies. For radical advocates of democracy, direct The second stage involved organizing democracy is an ideal and goal. Most of the personnel for the trial, including case world falls far short of this ideal. However, managers to represent both sides as well as even within systems of autocracy and repre- referees and judges. The third stage included a sentative government, it is possible to have range of things including case managers degrees of participation in decision making. developing factual statements, cross examina- What we call here “citizen participation” tion by opponents, and rulings on facts by refers to types of participation that occur judges (Cole 1986). within a system that is not fully participatory. Though much discussed, the science court Citizen participation can be considered both a has never eventuated. One of the purposes of good thing itself and, by some, as a training the science court was to sort out disagreements ground or staging post for moving towards about facts so that the differences over values direct democracy. would be more clearly articulated (rather than In this chapter we look at some examples of hiding in the midst of scientific controversy). methods of citizen participation that do not However, since the time the idea of the involve random selection. Each of these science court emerged, much work in the methods has strengths and limitations and social analysis of science has suggested that potentially can be improved through the use of attempting to separate facts from values is a random selection. We omit methods of futile exercise (Barnes 1974; Feyerabend consultation or participation that are one-way, 1975; Hesse 1974; Salomon 1973). In practice, involving very little discussion or debate, such every fact is bound up with values, due to the as public meetings where information is influence of sources of funding, professional disseminated or focus groups or surveys where vested interests, and assumptions embedded in information is solicited. systems of knowledge. This means that trying to separate adjudication of facts from discus- SCIENCE COURTS sions about values, and keeping citizens from being involved in discussions about facts, Science and society are often perceived as actually serves to give scientific experts a being worlds apart, with science dealing with privileged role in setting the agenda. In most Citizen participation without random selection 37

of the really contentious issues involving hensible information from witnesses alleviated science — such as controversies over forests, this. The balance of time between witnesses animal experimentation, weapons, chemicals, presenting evidence for either side of the issue urban planning, and genetic engineering — the was often unequal. “In the end the court’s issue of what counts as expertise and who is suggestions involved a compromise solution an expert is a basic part of the dispute. including careful monitoring, public disclosure Citizens do not expect to be involved in of activities, and broad participation by the deciding which theory of black holes or public in risk assessment. The city council radiative transfer is correct, but they do have a voted unanimously to accept the court’s role to play whenever science and technology recommendations” (Cole 1986: 253). impinge on their lives (Sclove 1995). Interestingly, the questions raised about the value of the citizen court could be applied to CITIZEN COURTS many methods of participatory decision making. Firstly, the manner of choosing lay External events overtook the science court persons remains undefined and problematic. idea and enthusiasm waned (Cole 1986: 250). Secondly, members who are uninformed but However, another model that bore many articulate might be unduly convincing. similarities did eventuate in Cambridge, Thirdly, allowing citizen courts to decide Massachusetts: the citizen court. This court about major issues can lead to inconsistencies was also meant to consider controversial between one panel and the next. scientific or technological issues but was less We can address the second difficulty with formal (relying on lay persons rather than robust group processes and skilled facilitation. judges) with the appointed citizens establish- Random selection is relevant to the first and ing their own procedures. The model still the third difficulties. Randomness eliminates emphasized a two-sided (for and against) many of the problems associated with approach, maintaining the court model but selecting lay persons for a citizen court without the cross-examination component (perceived bias, nepotism, unrepresentative- (Cole 1986: 252). A significant difference ness, and so on). between the two models lies in the contrast between the science court’s assumption that ADVISORY COMMITTEES technical matters can be considered in isolation from values and the citizen court’s Advisory committees are a standard form of belief that such a separation is not possible community participation in local government. (Krimsky 1986: 199-200). When a similar participatory method is used in Citizen courts have been convened in state or federal government, the committees Cambridge on two separate issues: chemical are often given grander titles such as advisory warfare research and recombinant DNA councils or standing committees. Advisory (rDNA), in both instances in reaction to the committees are seen as a way of soliciting siting of a laboratory within the city’s advice from the wider community to assist the boundaries. In relation to the latter, a panel of deliberations of elected representatives and/or eight citizens was appointed by the local administrators. The names of these committees council to offer advice, specifically on can vary: an economic think tank, an airport whether federal rDNA research guidelines advisory panel, a community consultation contained enough protection for the local forum, a public transport working group, a community (Cole 1986: 252). It has been floodplain management committee, and so on. suggested that such a panel could work in The quality of the “participation” can also tandem with a science court. vary. Committees can have the appearance of Inevitably, lay persons struggled with participation without this being translated into technical information and scientific argu- influence. ments, though a process of seeking compre- 38 Random selection in politics

In local government the levels of responsi- and lobby groups are usually higher in socio- bility of advisory committees vary as much as economic status than the people they claim to their names. For example, in Australia a sports represent (Gittell 1980). This raises important association committee might very well be issues of representation (that is, if representa- allocated its own budget that it can spend tiveness is to be the basis for selection of without constant reference to the local council. committee members). So too might a community center management This approach, however, is not used too committee which might be responsible for often. The more common method is for the tenancy agreements and oversight of mainte- governing body to designate or appoint a nance. Such committees are accountable, group of “experts” to come together for a few perhaps through annual reports, and the extent meetings to assist in the policy’s development. of their delegated authority is usually tightly If there is a broad consensus among those in controlled. However, it is not unusual for the the government entity, then it is most likely functions to involve a higher degree of that the expert advisory board will consist of autonomy than might be implied by the term experts who will be supportive of the “advisory committee”. This allocation of consensus. budget would not occur in the United States. There is a Catch 22 in this representa- Local authorities adopt a variety of tive/unrepresentative issue. If only specialists approaches in selecting advisory committee (those who are keenly interested and involved members. One method is to place a notice in already) are wanted for these committees, then the local newspaper and call for nominations. advertising for committee members in the Community members respond with a letter of local newspaper is probably not the best application. Those responding must see the strategy. Decision-making bodies need to be notice in the paper or be alerted by elected very clear about their goals and to determine representatives or council staff, and be literate the strategy most likely to meet these goals. and confident enough to apply. Inevitably, For example, if an agency is needing special- such committees attract the strongly interested ists, a better strategy could be to use a process and the highly educated — those who want to of contacting those are known to be interested change the way things are currently done and and involved, asking those persons who else those who enjoy spending time on a commit- would/could make a useful contribution, then tee, talking and making decisions. Meetings phoning this person and repeating the same for advisory committees are usually held questions. It takes very few phone calls until outside business hours so members are the same names are heard and the field can be inevitably drawn from those free to do so. This narrowed down to those who have the most to discourages single parents (usually women) offer. To achieve a range of views, potential and older people (who are often fearful of participants can also be asked “Who would being out at night). A portrait is therefore you suggest who would have an opposing emerging of the usual committee member: view?” educated, middle aged, and male. However, if representativeness, in terms of Lyn Carson’s analysis of one regional gender, age, occupation, education, and council in Australia showed this to be the case. geographical location, is important then There was a highly disproportionate number of strategies involving self selection are clearly middle aged or older, educated men. The inadequate. This is where random selection majority of the committee members were in could be applied. If a precinct committee, for full-time, paid employment; 70% of commit- example, was made up of randomly-selected tee members were senior managers, para people (one from each street or closely-linked professionals, or professionals (Carson 1996). streets), a cross-section of residents would be Just as most elected representatives have involved. This is particularly useful should a greater status and wealth than average voters, local council be interested in community so citizens involved in advisory committees opinion on planning issues. Citizen participation without random selection 39

Selecting randomly works less effectively, a similar position of power to the science of course, if the issues are ones where the courts discussed above (Joss & Durant 1995: incensed and the articulate have most to offer 17). The consensus conference, however, — for example in reaction to poor decisions opened the doors of the closed science court on planning. The participants in such an design, moving beyond a panel of elite partici- instance are acting as lobbyists within a pants, instead allowing for public input and flawed representative system of government attendance. and are indicative of an absence of timely The original concept for a consensus consultation. conference was developed by the United Issues of bias and conflict of interest States Institutes of Health in order to review inevitably emerge in advisory committees. In existing medical knowledge, disseminate this the regional case study mentioned above, the knowledge, and provide counsel on the use of conclusions of a floodplain management new technology. It was originally designed as committee were repeatedly dismissed by non- a two-step process. The first step involved the members who saw the committee as being establishment of agreement amongst experts made up of those who lived or conducted (about the introduction of new technology). businesses in flood-prone areas. They had The second step was to involve interaction most to gain from the construction of levees with experts in other areas (policy, planning, that would impact unfairly on those outside administration, ethics). This second step never the floodplain. A sports association in the eventuated. There have been over one hundred same regional city offered another example of medical consensus conferences conducted in perceived bias and conflict of interest. The the U.S. and the idea has also been taken up by association was often criticized as favoring its medical counterparts throughout Europe one sport over another, depending upon the (Joss & Durant 1995). make-up of the current committee. The Danish Board of Technology has The above advisory committees also reflect developed the model further, taking CCs a top-down approach to decision making with beyond the medical domain (Fixdal 1997: governing bodies permitting input from 375). The distinctive difference has been the citizens. A similar charge of elitism could be replacement of an expert panel by a panel of leveled against some residents action groups lay people who have access to experts. that emanate from citizens themselves. Such The Danish panel members are required to groups often replicate the same hierarchies have no in-depth knowledge in the technology they are keen to criticize, though this need not that is to be assessed (e.g., genetic engineer- be the case. However, representativeness even ing, food irradiation). In the early stages the in community-based groups is not necessarily panel members are given basic information considered a priority since fairness is not the about the topic under consideration. They basis of their formation. determine the agenda for the conference which runs over four days, two of which are open to CONSENSUS CONFERENCES the public. This freedom to set the agenda is a distinguishing characteristic of CCs; this Consensus conferences (CCs) may offer a significant self-managing element is often middle path between direct democracy (see absent from other consultative mechanisms. Chapter 3) and representative government, in During a CC, expert witnesses — selected relation to democratizing technology. CCs by the panel from a pool proposed by the reflect a participatory, deliberative approach steering committee — are called on the first that can influence policy making, participants, day of the conference. These experts, repre- and experts alike. Consensus conferences, senting a range of views and including both though an effective means of facilitating professionals and nonprofessionals, present decision making, have no authority other than information and respond to the panel’s to make recommendations. This places CCs in questions. The panel writes its own report and 40 Random selection in politics

recommendations. These recommendations are Board of Technology purposely selects in such publicized and passed on to elected represen- a way that lack of familiarity with the topic is tatives in the Danish Parliament. According to essential, random selection would result in Simon Joss (1998), parliamentarians who he participants who are at least as knowledgeable. surveyed showed a high awareness of the Indeed, the Danish Board of Technology process and saw it as having been quite has recently begun to use direct mail invita- influential in final decision making. But tions of randomly selected people from the Marcus Schmidt (personal communication, 17 Danish electoral register (Lars Klüver, February 1999), who has followed CCs personal communication, 9 December 1998). closely, doubts that any Danish CC has ever A British organizing committee plans to use had a significant impact on parliamentarians or randomness to select its lay panel for a CC on legislation. Despite the time and expense radioactive waste scheduled for May 1999 involved in planning and running a CC, it (Jane Palmer, personal communication, 10 costs less than alternatives such as a random November 1998). Other countries have been sample survey or a traditional technology reluctant so far to follow this lead. Both the assessment project using experts. Canadian and Australian steering committees Consensus conferences have been held in believed random selection to be more difficult, Austria, Britain, Denmark, France, Japan, time consuming, and expensive. Both Korea, Netherlands, New Zealand, Norway, organizing groups believed the sample which and the U.S., and are under way in Australia would result from using random selection and Canada. Consensus conferences can be would not necessarily result in a better cross considered to have two aims: to promote section of the community than one derived public debate and contribute to the making of from advertising for volunteers. political decisions (Joss 1998: 5). Indeed, an A counter argument is that random selec- important component of their success has been tion does make a difference, as suggested by media exposure of their deliberations and studies surveyed in Chapters 5 and 6. There findings. It has meant that the issues are have been no suggestions that criminal juries discussed more widely than the debate that be selected from volunteers, presumably occurs at the CC itself. because this would be seen to skew the The selection process is designed to create composition of juries, or at least appear to do a committee whose members reflect a broad so. cross section of community values, actually The CC process holds great promise designed to be more mixed in this sense than a beyond Denmark where it has become an representative committee. Usually 10-16 accepted method for allowing more voices to panelists are selected following their response be heard in the debate on technology. The to newspaper advertisements, submitting a Danish experience shows that technology has short description of themselves, the knowledge been controlled through the process. The lay they have of the topic, and their reason for panel’s recommendations that the irradiation participating. They are volunteers who of food should not be supported were used to typically are unpaid, though the Danish Board support the Danish negotiators’ position at the of Technology compensates them for loss of European Community negotiations over food income. The panel is selected to achieve a irradiation. Money was withheld from the comprehensive composition on the basis of a 1987-1990 biotechnological research program number of demographic criteria: age, gender, following a consensus conference on gene education, occupation, and area of residence. technology in industry and agriculture (Joss & Norwegian consensus conferences have Durant 1995: 128). In 1998, as a direct result included an element of random selection: once of a lay panel’s recommendations about responses have been collated, half the partici- mapping the human genome, the Danish pants are drawn from a hat and the other half government passed a law on the use of gene selected by the organizers. Since the Danish testing for employment and insurance Citizen participation without random selection 41

purposes (Lars Klüver, personal communica- by sex, age, location (as inferred from phone tion, 9 December 1998). However, a more numbers), or other variables. skeptical view (Marcus Schmidt, personal The strength of electronic hearings is a communication, 17 February 1999) is that broadening of the participation base, increas- government decisions have not been greatly ing public involvement beyond any numbers influenced by CCs, and that other influences that might be feasible for advisory committees or coincidence explain the correlation between or consensus conferences. Many people can CC recommendations and subsequent watch and listen to a debate. Many people can legislation. either comment on or vote on the subject Inevitably, the prevailing political climate matter. This input from a wider public can will influence the extent to which a lay panel’s clearly influence decision making. In the recommendations are adopted. A significant Honolulu example mentioned above, many of by-product of CCs is the influence that the lay those physically present in the chamber had a panel wields on experts whose horizons may vested interest in renovating the Waikiki Shell, be widened by the unexpected questions and being involved with the developer or labor perceptions on the panel. unions. Home viewers were far less impressed with the project, which lost the vote by a 3-1 ELECTRONIC HEARINGS margin (Becker 1998).

In the U.S. there have been instances in which CITIZENS PANELS public hearings have been televised with a participatory twist — citizens watching the The term “citizens panel” has been applied to televised hearing had the opportunity to testify a range of consultative approaches that can at the hearing by telephone. In one example in have different methods of selection. A citizens 1987, telephone callers were linked into the panel might be the term used to describe an Honolulu City Council’s internal public advisory committee (described above) or a address system when the Council was making forum or working party and would be either its decision about whether or not to renovate self-selecting or via appointment. Membership the Honolulu’s Waikiki Shell. Citizens who is usually defined and the group might meet were physically present and those calling from regularly (though not always). A forum might their homes waited their turn and were each be a once-only event with an open invitation given a brief (one minute) opportunity to and might be little more than a public meeting. testify. A panel would have a fixed membership and This method can be combined with an members could have been invited to partici- opportunity to vote: viewers watch the pate, to represent particular organizations. proceedings, then call a telephone number to Participants have an opportunity to debate vote on a proposal — for or against. issues that are defined by the commissioning Compared to the usual U.S. public hearings, in body which also makes the decision about electronic hearings access is broadened to what to do with the results (Coote & Lenaghan include more citizens. Electronic hearings also 1997: 7). have an advantage over a referendum in that, However, “citizens panel” is also the term before a vote is lodged, participants have been sometimes used to describe a “citizen jury” exposed to discussions that canvass a range of which uses different methods of recruitment opinions. The method of consultation is self but relies on random selection as one compo- selecting, neither universal nor random, nent. Often the random selection will be by though these could arguably be factored into stratified sampling, to match a profile of the the process. For example, if large numbers local population derived from census and other called in to speak, the sequence of callers data. For this reason citizens panels/juries will invited to comment could be decided by be taken up in the next chapter on citizen means of random selection, perhaps stratified participation with random selection. 42 Random selection in politics

CONCLUSION ultimately to citizens’ control, though it would be difficult to find working examples of this. Without necessarily challenging existing However, the range of options is worth political structures, the above consultative listing in more detail. The list of options we methods overcome some of the deficiencies of have offered (see Appendix) is not exhaustive unilateral decision making. These methods and relates only to those methods that might could be described as two-step approaches: be appropriate in a political decision-making step one is consultation and step two is or policy-making context. The options have decision making. They can serve a number of been categorized as Informing, Passive purposes. Consultation, Active Consultation, and Partnership to signify their level of genuine • They can get public officials out of a jam. involvement and decision-making power When an issue becomes too hot to handle, (Carson & Roache 1996). The list includes when opposition is building, a two-step those using random selection (e.g., citizens approach shifts responsibility onto others. panels, surveys, polls) as well as those with • They allow public officials to respond to the potential to do so (e.g., precinct commit- community demands for a greater say. tees, issues forums, strategic planning) and Using an extra consultative step means that those which do not (e.g., listening posts, street officials can reassure the community that corner meetings, public rallies). their power has been shared. Let’s turn now to those processes that use • They can lead to better decisions. By testing random selection. acceptability through a consultative process, public officials know that the wider community is more likely to get the decisions it wants. By throwing open the decision making process, more creative ideas are likely to emerge. • They stand a better chance of ownership. If community members have been part of a decision-making process, they are more likely to own the consequences of that decision, even if it’s not as good as was expected. • They lead to a more empowered community, particularly if the agenda is controlled by community members and the decisions made are enacted.

These consultative methods take many forms other than those mentioned. Those using random selection are outlined in the chapters that follow, but many consultative methods do not use random selection. Consultative methods can be tokenistic or even manipulat- ive of the general public. But at the other end of the scale they can also be an exercise in partnership between government and citizens even if random selection is not involved. In theory, participatory processes can lead 5 Citizen participation with random selection: the early days

“I know of no safe depository for the 3. It is forbidden for others to lobby them ultimate powers of the society but the because they serve in their own right, not as people themselves, and if we think them formal representatives. not enlightened enough to exercise that 4. They must seek consensus.” control with a wholesome discretion, the remedy is not to take it from them, but to (As noted in chapter 2, actual jury systems inform their discretion” (Thomas Jefferson, letter, 1820). may differ from this ideal. For example, in Britain property requirements for jurors were In this chapter we look at the early days of only removed in 1972 and majority (that is, citizen participation with random selection, in non-unanimous) verdicts were introduced in particular at the work of two men whose 1967.) names have become synonymous with this Coupled with these elements are the subject. Any experiments with citizens panels characteristics that would be most essential or policy juries will inevitably begin with an when conducting a consultation process. For acknowledgment of the work done by Ned example, Crosby et al. (1986) consider that Crosby and Peter Dienel beginning in the early there are six criteria for successful citizen 1970s. Also often noted is the move towards participation. They distinguish lobbying from workplace democracy that was occurring participation, noting that lobbying is based on around the same time. It is to these pioneering the belief “that a particular view is correct” and enduring efforts that we first turn. and that it is necessary to express that view to The history of these methods actually starts a public official (Crosby et al. 1986: 171). In in the late 1960s when a number of people contrast, citizen participation works in were independently arriving at similar reverse: a diverse group of people is informed conclusions about the ways in which citizen about a topic and then reaches a view about participation in political decision making that topic. Crosby et al.’s six criteria are: might be organized. Notable for its synchro- nicity were the separate but similar develop- “(1) the participants collectively should be ments that began in the U.S. and in Germany [typical of the population from which they in the early 1970s. Known as citizens panels, are drawn and] selected in a way that is not policy juries, or planning cells, they had a open to manipulation; number of identical components which are (2) the proceedings should promote effective outlined below. (For both a description and decision making; critique, see Renn et al. 1995.) (3) the proceedings should be fair; Given that many of the methods are (4) the process should be cost effective; modeled on criminal juries, let’s take a (5) the process should be flexible; moment to look at the current conception of a (6) the likelihood that recommendations of the jury. Emery (1989a: 132) says that, as a group will be followed should be high.” theoretical ideal, juries have these four elements in common : With these components in mind, let’s explore some experiments in citizen participation “1. All eligible adults have a duty to serve. using random selection. 2. They are selected by lot, not by ballot, and are not appointed. 44 Random selection in politics

CITIZENS JURIES, POLICY JURIES: tion and it is assumed that all have a right to UNITED STATES join in judgment. Participation in criminal juries is compul- Ned Crosby was a peace activist and social sory for those called, at least officially, though change agent in the 1960s working in the in practice quite a few are able to argue or poorer neighborhoods of Minneapolis, otherwise wriggle their way out of the Minnesota. While working on his political obligation. Jury duty is one of a small number science doctorate on social ethics he came up of things that the modern liberal state compels with the idea of a citizens jury and subse- citizens to do. The others are paying taxes, quently founded, in 1974, the Center for New undertaking military service, and, in a few Democratic Processes (renamed the Jefferson countries, voting. However, participation in Center in the late 1980s). Without any citizen juries is not compulsory, and this awareness of Dienel’s similar work being makes a big difference. Random selection undertaken in Germany, Crosby tested a must be from volunteers, which means that consultative method that was designed to special care is needed in selecting the sample. empower the participants. When Dienel and The Jefferson Center’s juries are conducted Crosby finally met in 1985 they were amazed most often on the request of public officials by the similarities between their two methods who are charged with making a decision on (Crosby forthcoming). Crosby’s citizens juries the issue being considered. Crosby and his — sometimes called policy juries — were colleagues randomly selected citizens who modeled on the existing jury system which he attend a series of meetings to learn about a thought was, in the United States, a familiar specific issue and to render their verdict on and trusted form of small group decision what should be done. The issues are usually making in public affairs. (Note that the tough, controversial, and complex. Jefferson Center has taken out a trademark on Jefferson Center’s citizens juries have been the name Citizens Jury.) sponsored by a range of organizations, The use of juries in courts of law is based including government departments. The first on a belief that an impartial panel of citizens juries dealt with issues such as agricultural drawn from a cross-section of the community impacts on water quality, organ transplants, is the best means for making very important and high school clinics to address teenage decisions — often about freedom or incar- pregnancy, AIDS, and other sexually trans- ceration, and, in some countries, death. The mitted diseases. presumption is not that each jury will include a To begin, a sizable jury pool is compiled by complete cross-section of the community, with survey technique. The survey involves an representatives of all economic, social, intensive telephone interview process in order religious, racial, political, and geographical to ensure that important demographic groups. Rather, this form of selection is based variables are proportionately present on each on a premise that prospective jurors are jury panel. Questions are asked to determine selected by court officials without “systematic age, sex, education, race, and geographical intentional exclusion of any of these groups” location. The telephone numbers are randomly (Bloomstein 1968 cited in Emery 1989a: 64). generated. Once a sufficient number of willing People are expected to serve on criminal people is found, random selection occurs until juries as a matter of civic responsibility, even specific categories are filled. From a randomly though in practice quite a few seek to avoid selected pool of 100 (sometimes more), a jury jury duty. Not all jurors will be intelligent; of 12-24 people is selected. The jury meets some will possess intuition or maturity of over a period of four to five days. Jurors are reasoning which does not rely on formal paid an average daily wage for attending. education; some will have little to offer in a Crosby has run these as once-only events or as debate; others will demonstrate leadership or a state-wide project with multiple juries being debating skills. This reflects the wider popula- Citizen participation with random selection: the early days 45

convened in different regional settings, One way to achieve the appropriate ratios is to culminating in a statewide panel. set up quotas for each category, such as To ensure that participants reflect a balance female, white, high school graduate, 55-64, of views in relation to the policy problem, the urban. Using random phone numbers, new Jefferson Center also stratifies the sample volunteers for the pool are assigned to their based on attitude to the subject matter, rather relevant categories until each category is filled than on a demographic basis alone (Crosby et up. In effect this was what the Jefferson al. 1986: 174). Center did, though the procedure in practice Random selection is central to the citizen was a little different. By using this method of jury, ensuring that bias is seen to be ruled out stratified sampling, the jury pools were an in selecting members, who are neither accurate reflection of the demographics of the appointed nor self selected. However, the state. voluntary nature of participation requires that In selecting the actual jurors randomly from random selection be carried out in an the jury pools, a further stratification was especially careful way. Consider, for example, used. All those in the sample were asked their the policy juries on school-based clinics for opinions about school-based clinics — the teenage pregnancy, AIDS, and other sexual topic of the juries. Then the jurors were transmitted diseases (Jefferson Center 1988). selected randomly so as to reflect the same The Jefferson Center ran eight policy juries, distribution of initial opinions. In this way, one in each of Minnesota’s eight congressional there could be no suggestion that the jurors districts. The initial approach to possible were a biased sample based on their prior jurors was by phone, to determine whether beliefs about the issue at hand. This example they would be willing to be in a jury pool. shows how random selection in jury construc- Randomly generated phone numbers were tion can be tightly controlled to ensure that used, taking into account relative populations randomness determines precisely who is in different area codes. Of all the people selected but that the jury overall is demog- approached by phone, about one third refused raphically and attitudinally typical of the to be interviewed at all. Those who agreed population. went into the jury pool. The Jefferson Center Once the jurors are selected, they are called finds that only about one out of five people in together for the actual process. The Jefferson jury pools actually agrees to join the jury. Center establishes the agenda for the jury and Since the Center wanted 12 members and four provides skilled facilitators to help the jury in alternates for each jury, it aimed at creating a its task. The facilitators take special care to jury pool of 100 for each congressional avoid being partisan, since this would district. undermine the credibility of the process. This The jury pools were constructed using is similar to the neutrality expected of stratified sampling. The aim was for members electoral officials who organize elections and of each pool to collectively match the collect, count, and report the vote. In a policy demographic characteristics of the entire adult jury, the agenda, the choice of speakers, and population. In practice, this meant seeking to the facilitation are all designed to avoid create pools with the appropriate ratios for the manipulation and instead to empower the following demographic characteristics: participants. • sex (female and male); The jury hears “witnesses” who present • race (white and non white); “evidence” or outline their positions on the • education (less than high school; high issue under consideration. The Jefferson school graduate; some college; college Center organizes proceedings to represent the graduate); major points of view in a fair and respectful • age (18-24; 25-34; 35-44; 45-54; 55-64; manner. The aim is to allow, via a set of 65 and over); structured hearings which represents all • urban and rural. 46 Random selection in politics

viewpoints, sufficient learning to occur so that consider the responses, and follow this up with an informed decision can take place. further questions and discussion. The After running the eight policy juries on weakness is that many people are unwilling or school-based clinics, a statewide policy jury of unable to devote five days to a policy jury. 24 was held, made up of three members Since Crosby’s jurors are paid, the conve- chosen from each district jury. Rather than ner must recognize the financial value in the selecting the members for the statewide jury exercise. Payment is, of course, a strength for randomly, they were chosen (elected from participants who are too often called upon to volunteers) by the district juries. Whereas the give up their time in processes which are so district juries had demographic characteristics often undervalued. closely similar to the state’s population, the The biggest hurdle that Crosby has had to statewide jury did not, having for example face is that politicians are seldom willing to fewer women, non whites, young people, and fund or use the method. Since the outcome of people who had not finished high school. This a jury cannot be predicted or controlled, it is illustrates how self selection and elections, quite a risky procedure for supporters of the even within the policy jury context, can lead to status quo to adopt. Frustration and idealism a choice of representatives who are not typical have occasionally led to its use (Crosby 1990: of the population. 35). Crosby notes that one of the strengths of the The biggest strength of citizen juries has process is that jurors “offer reasons for their been the response of participating jurors from decisions and point out what they like and all walks of life and all political persuasions. dislike about the alternatives” being consid- Participants have rated it very highly in terms ered (Crosby 1990: 37). One of the dangers, of enjoyment, fairness, and for generally being particularly for elected representatives who a valuable experience. The results indicate that might consider convening such a jury, is that citizen jurors perform extremely well and the outcome is not predictable and there must usually arrive at reasonable, thoughtful, and be an act of faith in the jurors’ ability to make widely acceptable solutions. The media also wise decisions. report favorably on the process and its results Again, there is a contrast with a criminal and some politicians have been similarly jury. The criminal justice system tightly praiseworthy. constrains the role of the jury. Police and Juries convened in the U. S. have arguably prosecutors make decisions about charges, and had little impact on decision-making bodies the judge and counsel make decisions about and are less often convened by them than they running the case. Jurors are not allowed to ask are by enthusiastic practitioners such as questions during the case and sometimes not Crosby even though they have attracted the even allowed to take notes. The criminal jury favorable attention of the media. Crosby makes a judgment within tight bounds, which believes that the policy jury method will gain often boils down to a choice between guilty significant power only when the public trusts and not guilty. In contrast, policy juries have in the process and demands its use (Crosby much greater scope to question witnesses, 1990: 35). propose alternatives, and develop their own recommendations covering a range of matters. PLANNING CELLS: GERMANY With such a relatively open-ended process, it is not surprising that politicians are wary of Around 1970, Peter Dienel from the Univer- giving too much power to citizen juries. sity of Wuppertal in West Germany developed The four to five day length of Crosby’s jury the idea of the “planungszelle” or planning process is both a strength and a weakness. cell. In 1973 he established the Institute for Devoting so much time increases the possibil- Citizen Participation and Planning Methods ity for genuine deliberations to occur. Jurors and began experimenting with dialogue in can take time to reflect, ask questions, small groups to improve governmental Citizen participation with random selection: the early days 47

decision making (Crosby forthcoming). He did population is composed of factions with this by using randomly selected groups of different interests. citizens to deal with issues ranging from city Determining who is affected also has its planning to future energy paths. Planning cells problems. People might be affected by the normally are groups of about 25 people who construction of a multi-lane highway that work together for two to five days on could destroy fauna and flora and lead to problems of planning, assessment, or control increased greenhouse emissions even though (Dienel 1988, 1989; Dienel & Renn 1995; they do not live in the construction area. The Garbe 1986; Renn 1986; Renn et al. 1984). next generation, too, may suffer the effects of During this time, the participants are consid- current decisions. Their voices cannot be ered to work as public consultants. Their heard. A rehabilitated industrial zone that is living expenses and wages at their usual jobs being converted to a housing estate has no are paid for either by government or employ- affected residents with whom to consult — ers. Dienel has worked with single groups as yet. So who would make the decisions in these well as simultaneous groups dealing with scenarios? Dienel believed that these national issues (for example, seven groups in anomalies are best addressed by giving every different cities). citizen an equal opportunity to serve the wider Dienel’s work was meant to create a new community and that this equal opportunity is model of political involvement for German best achieved via random selection. citizens, affirming their role as being Dienel also believed in equally sharing the sovereign in any democratic political system. complex task of decision making. No single Dienel was clear that he was not speaking individual need play the role of universal about the “professional citizen” when citizen (a role which elected representatives discussing empowerment. His interest was less are asked to play). In order to involve many in elevating or maintaining the role of those citizens in decision making on the same issue citizens who might already be publicly vocal in a short period of time, Dienel proposed because of their interest in a political career, organizing simultaneous planning cells personal benefits, or public image (Renn et al. (Dienel & Renn 1995: 121). Citizens should 1995: 120). His quest was to find ways in be asked to serve for a limited time only. which virtually anyone could play the function These features of random selection, temporary of a decision maker if his or her life was assignments, parallel problem solving, and affected by the decisions. division of labor have been combined in the This latter motivation — that citizens planning cell model. should have a right to participate in decision Planning cell participants are exposed to a making which affects them — is more range of information in different formats: via contentious than it might first appear. We have hearings, lectures, panel discussions, video- already noted that, in representative govern- tapes, and field tours. The idea is again ment, arguably that right has been forfeited, modeled on a jury trial with experts and given that representatives make decisions on stakeholders as witnesses. Most of the task behalf of the population. One might delve work by the planning cell is completed in further into this and ask, “Should democracy small groups of five, a process that encourages be based on affectedness?” (Seiler 1995: 146). everyone to take an active role and enhances That is, should participation be linked to the the possibility of creative solutions. (Dienel extent to which one is affected? Hans-Jorg and Renn 1995: 122). Seiler, writing from the viewpoint of a Swiss The table below was developed by Dienel lawyer, points out that such a notion is often and Renn (1995: 123) to describe the structure spoken of pejoratively in European countries. and conditions that are necessary for planning The European conception of democracy cells. emphasizes the idea of a common interest, in contrast to the typical idea in the U.S. that a 48 Random selection in politics

Structure Condition Step Four Introduction of options through lectures (designed Composition Random selection of directly and to be non-controversial) or hearings (designed to indirectly affected citizens. highlight areas of disagreement). Involvement of stakeholders and public officials as witnesses, not Step Five as participants. Problem-structuring with respect to each option through group sessions and plenary discussions. Task Evaluation of different decision options in accordance with Step Six personal values and preferences. Evaluation of options through the use of individual Clear political mandate to draft questionnaires and group discussions (captured in recommendations for legal group response forms). decision maker. Step Seven Operation Full-time meeting over several Drafting of rough recommendations through work- days. groups and plenary sessions. Receiving information about likely consequences of each Step Eight option. Articulation of recommendations in a citizens’ Incorporation of uncertainty and report done by the facilitator after the completion dissent through public hearings of the planning cells. and videotapes. Step Nine Roles of Identification of participants as Feedback of citizens’ report to participants participants “value consultants”. (usually in an evening meeting two months after Need for external, neutral, and the planning cells). unbiased facilitator. Low involvement of sponsor Step Ten (confined to witness role similar Presentation of the citizens’ report to the sponsor, to stakeholders). the media, and interested groups.

Organization Payment of an honorarium to There is a subtle but clear difference in the each participant for working as value consultant. role of stakeholders, those who are interested Local organization committee for in or affected by a decision, in the German and facilitating the invitational U.S. models. According to Dienel and Renn process. (1995: 127), in the U.S., “stakeholders have a long tradition of being included in decision Dienel and Renn have also documented the making,” at least in theory. This is exemplified sequenced activities which occur in a typical by the commitment to public hearings prior to planning cell. These steps are similar to those legislation. In practice those involved are less adopted by other practitioners. Dienel and likely to be citizens and more likely to be Renn (1995: 124) present them as follows. those with a financial “stake” in proceedings. In Europe it is more typical for stakeholder Step One participation not to be legally mandated or Introduction to the issue through lecture(s) and ensured through tradition. In both places there field tour. is tension involved in the degree of Step Two stakeholder participation. A balancing act is Provision of background knowledge through performed in involving stakeholders or lectures, written materials, self-educating group sponsors in the review of information sessions, audio-visual information, field tours, and materials, suggesting experts, and so on. A others. line is drawn to avoid any influence that might Step Three be wielded on the panel or cell, or on the Introduction of conflicting interpretations of recommendations that emanate from either. information through written materials, videos, Planning cells have been used not just in hearings, panel discussions. Germany but also in the Netherlands, Citizen participation with random selection: the early days 49

Palestine, Spain, and Switzerland (Coote & The German experience of planning cells is Lenaghan 1997: 13). The Swiss experience an inspiring one. Organizers point out the provides an example of the culturally-specific significance of involving those who are not in nature of random selection. In Switzerland, paid work in planning cells; they see it as a instead of using random selection, the way of sharing problem solving and public oversight committee recruited and selected service. Participants report that the process participants. The sponsoring agency was helped them become better communicators concerned “that random selection would not and made them more confident in creating be seen as a legitimate way of choosing cooperative working styles. They recognized representatives”, given that “lotteries as a their role as consultant or advocate for the political means of achieving equity [are] alien public good. Planning cells are recognized as a to the Swiss political culture” (Dienel & Renn legitimate component of government decision 1995: 133). The Swiss, of course, have a making and are usually commissioned by the tradition of heightened public involvement in authorities which have power to act on their decision making through their cantons and recommendations. community councils. A similar resistance to random selection was expressed to one of this WORKPLACE DEMOCRACY: book’s authors when organizers in Denmark AUSTRALIA were questioned about the absence of random selection in Danish consultative processes. We have been talking thus far about random Within the Danish tradition of participation, selection being used in community planning or self selection is seen as acceptable. A tradition problem solving. Random selection is an of participation is compatible with use of adaptable technique and the possibility exists random selection but lotteries are more likely for its transference into the workplace. In the to be associated with fairness in some creation of industrial democracy, might countries than others. workers’ names be drawn at random and Other cultural differences emerge when a worker members become part of a depart- process is transplanted across continents and it mental committee that involves itself in is clear that one needs to adapt a consultative departmental planning and co-ordination technique to the affected community. decisions? According to Dienel and Renn, in West Beginning in the 1950s at the Tavistock Germany participants were apparently Institute in London, researchers including Fred “grateful and pleasantly surprised” that Emery, Einar Thorsrud, and Eric Trist became someone made the effort to plan and structure involved in studying and promoting democ- a participatory procedure whereas in the U.S., racy at work (e.g. Emery & Thorsrud 1976). “citizens distrust pre-fabricated participation They observed that some groups of workers, models and suspect hidden agendas” (Dienel for example in underground coal mines, & Renn 1995: 136). However, Crosby themselves decided how best to organize the (personal communication, 21 January 1999) work. Such worker-managed systems brought reports that his experience is opposite to this. out the best in workers and were quite Residents of Denmark, according to a efficient. This form of direct democracy at number of surveys, are far more knowledge- work inspired the researchers to see how to able and better informed than their counter- promote “sociotechnical design,” in which parts in other European countries. Further, the workers helped to design participatory work notion of trying to achieve consensus has been systems. The idea was that technology and surely nurtured in a country such as Denmark work practices are chosen by all concerned — where no single political party held a majority management, workers, and labor unions — to in the Danish Parliament for most of the 1900s achieve commonly agreed objectives. — open debate and negotiation is a political In any process of designing work, key way of life (Toft 1996: 174). questions are who makes the decisions and 50 Random selection in politics

how. In traditional bureaucratic structures, view. Finally, contributions were invited from decisions are made by top managers, often by the entire membership. Through use of these the chief executive officer, with or without methods, input was obtained from a wide cross consultation. When large numbers of workers section of members. Some of the members’ participate, a wider range of options becomes cynicism about the union was countered when possible, including consensus, voting in face- they realized that the consultation was genuine to-face meetings, election of representatives, (Davies 1982). and sortition. Fred Emery in particular was The discussion groups in these extended attracted to sortition for workplace decision consultation exercises had variable attendance. making. He observed random selection used in For some groups everyone invited showed up, three workplaces: a fertilizer plant in Norway, whereas in others no one at all appeared! The a heavy electrical manufacturer in India, and a most important factor was personal contact: small high-technology company in Australia when the group convener personally contacted (Emery 1989a: 74). Workers and managers in members, face-to-face or by phone, they were each case decided on random selection after far more likely to attend. considering alternatives. In his lifelong pursuit of participatory In the 1980s, fellow Australian Alan Davies democracy, Emery became ever more suppor- introduced some of Emery’s ideas into the tive of sortition (Emery 1989a). He was highly trade union movement in Australia. Davies critical of representative systems, including worked with the Tasmanian Public Service workplace councils with employee representa- Union (TPSU) to help it restructure to become tives, a typical model for “industrial more effective (Davies 1985). As the first democracy.” By choosing members of phase of a strategy for change, he conducted a workplace councils by lot, there is less risk “search conference” (see Emery and Purser that those selected will become tools of 1996) of 42 invited participants. In the second management and isolated from their fellow stage, 600 TPSU members were chosen workers. Emery (1989b) also advocated lot- randomly from the union membership of based selection of various policy advice 10,000. The 600 were put into 85 discussion bodies, such as industry councils that are groups, whose meetings were convened by currently appointed by government. fellow unionists who were trained in facilita- The democratization of work is of crucial tion. Davies described the randomly-selected importance, given that work is where most participants as a “deep slice” because the people spend the greatest amount of their time. groups included members from all levels of Decisions at work immediately affect the the organizational hierarchy. The deep slice quality of working life, not to mention wider using selection by lot provided genuine input impacts on productivity, social welfare, and from rank-and-file members (Davies 1985). the environment. When working environments Davies applied the same approach in are “rational,” namely decided at the top working with the Administrative and Clerical without worker participation, human capaci- Officers’ Association (ACOA), a major ties are imprisoned in what Max Weber Australian labor union covering government described as the “iron cage” of bureaucracy. employees, with a membership of 50,000. In George Ritzer (1996) has used the word order to work out how to better organize and “McDonaldization” to describe the change in run the ACOA in a changing environment, workplaces towards the fast food characteris- four different methods were used to find out tics of efficiency, quantification, predictabil- views of ACOA members. First was a search ity, and control. There seems little room for conference with executive members of the shared decision making or agenda setting union. Then there were some 300 discussion within such dehumanized work processes. groups involving 2000 members chosen by lot. Third, contributions were invited from members with special knowledge or points of Citizen participation with random selection: the early days 51

CONCLUSION comprehensible, that participants should be facilitated to complete their task, that the The name used by those who convene a policy participating group should not be too large, jury, a citizen jury, or planning cell is signifi- and opportunity for equal participation must cant. Language carries important messages be incorporated in the process. Procedures and the term “jury” is an especially powerful should be fair and seen to be fair. One needs to word, resonating with legal proof, grave deci- be careful about the sharing of information. If sions made by peers based on comprehensive information is being doled out, then the evidence. Colin Brown asks whether the same agenda is beyond the control of participants. reactions might occur to a consultative process They must be free to set their own agenda and that was called an “advisory panel” or a call on any additional information once the “workshop”? Probably not. We need to be topic has been canvassed. mindful that a citizen jury differs in a number Any consultative process runs the risk of of significant ways from a legal jury (Brown being an empty ritual when no clear agreement 1997). A legal jury deals with evidence that is is made prior to the consultation that the strictly determined by the court, delivering a decisions reached will be adhered to. Tentative verdict that is decisive. A citizen jury can call work has been done in this area, in asking any evidence it wants but its decisions to date sponsoring authorities to sign a contract, so have been merely recommendations. that participants are clear about the extent to Are citizens better at making policy which their final decisions will be enacted. decisions than politicians or bureaucrats? Participants are less likely to feel let down if Using citizens certainly minimizes the role of they know in advance the limits of the process. vested interests. But better at making the Authorities are also less likely to back down if decisions? Perhaps not. This is not a problem they have committed themselves to adhering of complexity. Ordinary citizens equipped to a decision reached by citizens. with accurate and meaningful information and Juries can attend to short-term problem time in which to debate and deliberate will be solving or long-term planning, local or equally effective in making decisions as those national. Local juries can be convened which currently making decisions on public policy tie into a national configuration. Evaluations matters. However, the same constraints that indicate that jurors are very satisfied with the exist for those in authority will rear up to process but little is known about the wider obstruct empowered citizens in their efforts to community’s knowledge or satisfaction with make “good” decisions. Constraints to it. Since recommendations to date have rarely effective decision making include those that been formally adopted by agencies, the are structural, systemic, psychological, challenge would seem to be to match the physical, cognitive, and interpersonal (Carson community’s dissatisfaction with current 1996). consultative practices by heightening their Random selection is both a way to limit awareness of alternative processes. poor quality decisions as well as a means to The three key groups influencing policy are create a more robust process that might elected representatives, interest groups, and remove some of the more obvious impedi- bureaucrats. This system may be entrenched ments to effective decision making. For but is not irreversible. How might decision example, random selection might mean that making by an elite minority be converted into there is less fear attached to the decision. a shared process? This chapter began to There is no power base to be maintained, no answer this question. The chapters that follow career on the line, no electoral ramifications, will explore the question further by looking at no allegiances to abide by, no entrenched some practical experiments that have built on patterns of faulty institutional decision making the work of Crosby and Dienel. to be perpetuated. It stands to reason that information should be written in a way that is

6 Citizen participation with random selection: yesterday, today, and tomorrow

In this chapter we will look at more recent that it trial at least one randomly-selected examples of citizen participation in political committee. She agreed to co-ordinate both the decision making which use random selection. selection and establishment of the Girard’s These different methods have been trialed Hill Precinct Committee. throughout in the world. We survey experi- Girard’s Hill is a small precinct with only ences from Australia, the U.S., and Europe 25 streets so it was considered appropriate to and comment on the methods’ effectiveness. have one representative from each street. The various methods have been categorized by Street numbers were randomly selected the location of the individual or organization (literally drawn from a hat by the Mayor). which created or trialed them. Each street had four numbers drawn and these were labeled first, second, third, and fourth COMMUNITY CASE STUDIES IN preference. It was assumed that there might be RANDOM SELECTION: AUSTRALIA difficulty getting residents to participate so it was important to have back-ups should the One of the authors, Lyn Carson, had heard of first-drawn participant decline to attend. the experiments in participatory democracy The CCC distributed leaflets to mail boxes conducted by Crosby and Dienel that were prior to the “lottery” to alert community described in Chapter 5 and wanted to trial members to the possibility that their names them in Australia. A number of opportunities might be drawn. This leafleting was also arose which allowed her to do this. As an considered to be of educational importance as elected representative on a local council she it drew the community’s attention to the was able to convene a randomly-selected consultative process that was under way. precinct committee in 1992. As a community Those drawn from the hat as first preferences representative on a school council in 1993 she were notified by mail of their selection. was able to convene a parent jury. In the Carson and a colleague followed this up by following year, as an academic and consultant, contacting each selected participant by phone. she was asked to conduct a randomly selected Where the name on the list was an absent policy jury in a neighboring town to consider a landlord, the tenant was visited personally. main street planning issue. These three The aim was to involve not just the owners but examples of participation using random also the residents who actually lived in each selection will be outlined. street. Carson and her colleague assumed that Randomly Selected Precinct Committee people might be reluctant to represent a larger geographical area but would feel sufficiently In 1992 a Community Consultation Commit- connected to and responsible for their own tee (CCC) was formed by Lismore City street that they would agree to attend. The Council. Lismore is in eastern Australia, in letter that was sent had emphasized the northern New South Wales not far from the importance of selected participants represent- coast. Among the CCC’s tasks was to establish ing their street in Council decision making. and evaluate the effectiveness of precinct The area had a higher than average propor- committees in the local area. Three precinct tion of older women. These women were committees were to be trialed, two urban and reluctant to come out at night to attend a one rural. Carson recommended to the CCC meeting or had reservations because of their Citizen participation with random selection: yesterday, today, and tomorrow 53

failing health. Overwhelmingly they suggested ness to participate. They have committed younger, male neighbors who they believed themselves before pre-selection so there is a would have much more to contribute. Most of high turn-up rate. In the Australian case those contacted seemed quite surprised to be studies, the participants had to be persuaded to asked for their views and considered that attend once their name had been randomly special expertise must be necessary in order to selected. The result may be a distinctively contribute anything meaningful to the different cohort since the disempowered, the discussion. less confident, and the less articulate will After Carson patiently explained that they inevitably defer to others. Possibilities of were the very people who were wanted, people providing opportunities for all citizens may be who knew what was needed to make their lost if no attention is given encouraging all street and their community a better place to voices to be heard. live, many were willing to “give it a go” (try it). Those who remained unwilling were Parent Jury replaced by the next randomly selected person. A couple of people were extremely excited The second jury that Carson conducted was about “winning” the position on the committee held in 1993 (Carson 1994). Lismore Heights and could not wait to be involved. Virtually all Primary School, like all schools in the state of of the men who were contacted agreed to New South Wales, had a school council that participate. was seen by the state government to be a way On the first evening only 14 of the expected of devolving management responsibility to 27 arrived. Tremendous cynicism was regions and then to each school community. expressed by some participants about the value The school council replicated traditional of the project and the likelihood that Council committee structures with formal roles would either listen to them or act on their assigned to office bearers, formal meeting concerns. As it transpired this was quite an procedures, and the exercise of considerable appropriate concern because Council (i.e. control over the agenda by the school seven of the 12 councillors), decided to “pull principal. the plug” on the whole project within weeks of The existing parents association was poorly the first meeting. It did this by withdrawing attended and it was difficult to gauge a cross the very minimal funding that had been section of parents’ views about the school. allocated to a staff person to service this This school council therefore agreed to hold a project as a part-time precinct committee parent jury that Carson was to co-ordinate in coordinator. an effort to involve parents in setting the This experience highlights the reluctance of school’s goals. people, particularly women (and even more Advance publicity was sent to all parents in particularly, older women) to be involved in the form of an illustrated announcement that community decision making. It demonstrates their names had been included in an important the cynicism of participants towards elected lottery. All students’ names were placed in a representatives and government bureaucrats. It large container and two names were chosen demonstrates the hard work needed to follow for each school year: kindergarten to year six. up a selection process to convince people to be An equal number of male and female parents involved. were designated. These parents were contacted There was a distinct difference in the selec- and told that they had won the “lottery” and tion procedure used for these Australian invited to attend an evening’s discussion about projects and those used by the Jefferson the school’s goals. Again, participants were Center. The latter compiles a pool of willing followed up by phone. participants, derived using random selection. A fairly reluctant group arrived on the night Potential participants have been informed of the parent jury. Some had never visited the about the jury and have expressed a willing- school before. Some said that they were far 54 Random selection in politics

too ignorant to discuss such important issues. unfortunately misinterpreted some of the A skilled facilitator had been employed so comments. Organizers had been very sensitive various opportunities were given to parents to to the parents' needs but had stopped short of discuss their discomfort and reluctance. Many extending the same sensitivity towards the said they attended either from curiosity or teachers. guilt. Participating parents gradually began to In the following year, with a new acting feel as ease and realized that they were as able principal at the helm, a different process was as the next person to discuss one of the most used. A “think-tank” was convened and all important issues in their lives: how their parents were invited along to discuss the children should be educated. strategic plan. The usual handful of interested This gathering worked like a jury to the parents accepted the invitation and a signifi- extent that evidence was presented by a cant new process, the parent jury, has not been teacher, a counselor, several students, and a conducted again. community member. Each speaker spoke about their educational priorities and all Neighborhood Policy Jury avoided the use of jargon. There was wonder- ful, lively discussion from the parent group In 1994 Carson was asked to undertake a whose level of empowerment appeared to rise project for the Ballina Information Service (a by the minute. neighborhood center run by a paid coordinator Parents participated in small groups and the and a number of volunteers). Ballina is on the results, including many practical suggestions, east coast of Australia, near Lismore. The were enthusiastically reported back. The project was called “Our Ballina” and was parents, who had arrived reluctantly three designed as a visioning exercise to gauge hours earlier, turned into a forceful group and community opinions and ideas on develop- it was hard to send them home at the end of ment of Ballina’s central business district the evening, so animated and excited was the (CBD). discussion. With the help of a colleague, Carson trained Throughout the evening these parents three volunteers and the coordinator to plan realized a number of things and these realiza- and run their own policy jury. After a full- tions emerged in the evaluation at the end of day's training session and a little guidance they the evening: they were competent to discuss wrote all their own publicity material, negoti- these issues; some other parents shared their ated with speakers and Council, organized the views and some opposed them; all viewpoints venue and catering, gathered together were valid; discussions about education could materials for displays, and did a major mail- be stimulating and fun; expressing their views box drop throughout the area they wished to in this way could affect their children's future target. The training sessions took place in teaching and learning. February 1994 culminating in their neighbor- A significant blunder was made which is hood consultation six months later. worth noting. A commitment was made to Volunteers had to work hard to convince parents to pass on their comments and residents to attend. They had done a promo- concerns to the teaching staff which was done tional mail-box drop prior to the “draw” and via the teacher who had attended as a speaker. followed this up with a personal visit to those The list of priorities was not received well by who had been randomly selected. Older staff who interpreted many of the comments as women were particularly reluctant and criticisms and felt that they were doing many believed they had little to contribute. On the things already which were seen as important day of the consultation, the expected twenty by these parents. With the wisdom of participants was reduced to fifteen and the hindsight, it was clear that the facilitator majority were older residents (six women and should have been given the added responsibil- nine men). This accurately reflected the ageing ity of debriefing with teachers who had Citizen participation with random selection: yesterday, today, and tomorrow 55

population of Ballina, although men were retained and five things they would like to overrepresented. change. After reporting this to the large group The day began with introductions after they voted to determine the large group's participants had had a chance to look at the priorities. Volunteers from the Ballina visual displays, designed to stimulate their Information Service agreed to compile a report interest in the possibilities for developing the for distribution to Ballina Council, the CBD. An icebreaker followed during which participants, and the media. participants were prompted to think of a town In the weeks that followed, a number of other than Ballina that made them feel good participants called by the Ballina Information about being there and another town where they Service to chat and to see how the finished felt reluctant to leave. report was going. They were interested to see A variety of speakers, chosen by the neigh- what would become of the exercise. The borhood center for their variety of views, reluctant participants gave volunteers the offered their own perspectives. Those who feedback that the day had been a successful spoke included a town planner, a lecturer in one for them and that they found the process planning, an employee of the tourist informa- enjoyable. tion center, a member of the local environment The project was a good example of society, a high school student, and a council- grassroots participation. The idea came from a lor. Participants were given a couple of neighborhood group, was organized and run minutes after each speaker to discuss with by the group, and the final report was used to each other any questions they might wish to lobby the local government to act on the raise at the end or any points which sounded group’s recommendations. Four years later, interesting. This worked well. It can be the redevelopment of the CBD is being widely overwhelming to hear a string of speakers discussed though it’s difficult to assess the without interruption and it can also be extent to which the neighborhood group’s disruptive to have questions asked prematurely activities have pushed this agenda. which may be answered by later speakers. Questions followed and residents took this CITIZEN SURVEY PANEL: UNITED opportunity to make their own statements STATES about issues of concern. It had been suggested to the organizers that Though modeled on Crosby’s citizen jury, there be a brief exercise involving relaxation Boulder, Colorado’s Citizen Survey Panel and visualization, followed by small group (CSP) ran for an entire year. The panel’s issue work with clay, crayons, and craft to create was the transportation system of Boulder, a models. The organizers were decidedly contentious subject. Some residents and resistant to having older people being asked to planners favored an extensive highway system do these “undignified” things but reluctantly whereas others advocated facilities for agreed. This was a most successful session cyclists, pedestrians, and mass transit riders. with participants standing back watching until The Transportation Panel Project was a few people began to draw or create models, directed by Lyn Kathlene who recognized the then they started offering suggestions, finally importance of involving citizens at the policy working with the materials themselves. The formulation stage. The CSP was supported by group was very animated and came up with the Transportation Advisory Committee (made some innovative suggestions during this up of seven voting members) that was charged process. It was difficult to get them to stop for with creating the city’s Transportation Master lunch. Plan. After lunch the small groups reported back Random sampling was a key feature of the on the key elements from their groups, then CSP’s formation. Seven hundred residents discussed and agreed upon a list of five things were selected randomly, stratified by location they valued about Ballina that should be and housing type. Each resident was mailed an 56 Random selection in politics

invitation by the mayor to participate in the decision makers but it is a good example of a project (Kathlene & Martin 1991: 50). panel’s usefulness in affecting the direction of Responses were received from one third of the policy making. sample and a panel of 147 participants was To run the CSP, a policy analyst and a team finally formed. Participants were more highly of interviewers were required to develop educated and had a higher income than the information materials, design and administer general population though other important the surveys, collect and analyze data, and demographic factors were comparable. write up and present findings to the TAC. The According to the organizers, this was partly CSP used survey and interview techniques because of the required commitment to spend over a long period with a random sample of a year on the project. Thus home owners, who the city’s population, providing a more typical were less mobile, were more likely to partici- cross section of the Boulder community than pate. Nevertheless, later comparative surveys an open public hearing or an appointed showed participants to be similar to the wider advisory committee. Organizers were community in their assessment and use of city surprised and pleased that citizens were so services. willing to make such a substantial time A range of panel exercises ensued: four commitment to the panel, including reading surveys involving mailed questionnaires, one background information, being interviewed, survey by phone, and two interviews held at and filling out long written surveys. Few of home. The city paid participants to use the bus those involved had previously attended public system and report on the experience. To meetings; afterwards, most said they were maintain continuity, each panel member was interested in becoming more active in policy assigned to one of the 10 interviewers. Only issues. one out of 10 participants dropped out during Given its duration and scale, a panel such the course of the panel’s one-year life. The as the CSP is suitable for dealing with major panel provided the Transportation Advisory projects, especially in new policy areas where Committee (TAC) with considerable input in a the community’s views are unknown. It useable form. The data that was gathered from appears to have a number of advantages over this large panel was used by the TAC in four conventional forms of consultation, such as different ways. public hearings: (1) “Information that supported its policies • the participants are a more genuine cross and projects was readily held up as proof of section of the population; community support.” • the long duration of the panel allows great (2) Issues that divided the committee but scope for deliberation and reflection; had a clear community (panel) consensus were • the use of a variety of consultative tech- sometimes used to resolve a stalemate in the niques adds confidence to the findings; committee. • the credibility of the panel means that its (3) When the panel opinion differed from recommendations are more likely to be taken TAC opinion, this encouraged it to ax or seriously; revamp policies. • participants report favorably on the (4) Panel opinion that differed from the experience. TAC forced it to justify its policies, which sometimes meant rethinking the approach DELIBERATIVE POLLS AND (Kathlene & Martin 1991: 54). TELEVOTE: UNITED STATES The TAC’s radical recommendations were able to be supported by strong community Conventional opinion polls have a number of opinion (via the panel) and when announced deficiencies. In particular, responses may be received widespread support from the wider based on little information or consideration of community. The panel did not directly make the issue (Barber 1992). In contrast, “delibera- recommendations that were taken up by tive polls” are designed to poll the views of Citizen participation with random selection: yesterday, today, and tomorrow 57

typical citizens after they have been involved in at least two countries: the U.S. (Hawaii and in considerable discussion of the issues. In a Los Angeles) and New Zealand. These large-scale deliberative poll, a national random experiments, designed to create an innovative sample of citizens, several hundred in total, is system of “civic communication”, have been brought together for a few days. They are documented in depth (Slaton 1992). Unlike the given briefing materials in advance. When deliberative poll, televoting does not take together, they discuss the issues in small citizens out of their homes to a single site for groups and have an opportunity to hear and group discussion. Participants in televoting question experts and politicians. Participants sometimes are purely self-selected, but it is have expenses paid and may receive an quite possible to introduce random selection. honorarium and support to attend, for example A straightforward way to obtain a cross with child care and getting time off work. section of the community is to call randomly Their views are polled before and after the generated telephone numbers. If respondents event. The idea of the deliberative poll is to agree to participate, they are mailed informa- demonstrate “what the public would think, if it tion on the public issue to be discussed. This had a more adequate chance to think about the material includes opinions from various questions at hand” (Fishkin 1991: 1). The experts and a discussion of alternatives. deliberative poll, as its name suggests, is like Participants are encouraged to read the an opinion poll with the key extra dimension information and discuss it with family and of deliberation. friends. Deliberative polls were developed by James Televoting can be part of a wider process of S. Fishkin (1991, 1997). The first such poll involving as many people as possible in took place in Britain in 1994 (Fishkin 1996) discussion of issues. For example, the 1982 and was televised by Channel Four. In January Honolulu Electronic Town Meeting included a 1996, Fishkin organized the National Issues televote (with a random sample of 700), Convention, bringing 459 randomly selected newspaper articles, radio and television citizens from across the U.S. together in programs. To engage the audience, some of Austin, Texas to discuss the issue areas of the the television programs involved dramatic and family, the economy, and foreign affairs satirical commentaries with singers and (Fishkin 1997: 177-203). His aim was to use dancers. All the programs allowed for listeners the event, especially through associated media and viewers to phone in their comments. Thus, coverage, to exert a strong influence on public the televote was the more “scientific” dialogue before the presidential primary component of a larger process of promoting campaigns. A number of other deliberative citizen engagement with the issues (Dator polls have been organized in Britain and the 1983). U.S. Conventional opinion pollsters have Participants in televoting report that they analyzed, criticized, and debated the validity are highly satisfied with the process. Televotes and value of deliberative polls (Denver et al. have been used to help government agencies 1995; Merkle 1996). Deliberative polls can be make decisions on planning and policy issues. especially useful in dealing with topics about They can also be used to predict future voting which most citizens know very little, such as patterns accurately (Becker 1998). Christa local planning issues. They may be less useful Daryl Slaton (1992), who has been involved in on some general questions where opinions are organizing and documenting televoting not changed much by deliberation and where projects in Hawaii, Los Angeles, and New views can be measured more easily and Zealand, recognizes the value of collecting cheaply by conventional opinion polls. opinions via televote but also thinks that it is Like deliberative polls, televote is a method important to allow people to set the agenda. of polling public opinion. It is also known as The deliberative poll assumes that it is educational public opinion polling. Televotes valuable to bring people face-to-face for began in the late 1970s and have been trialed deliberation. Televoting allows people to 58 Random selection in politics

remain in their homes, though it can be readily citizen jury’s ability to deliver on these combined with electronic hearings, described characteristics. The table below is a slightly in Chapter 4, in which participants watch a simplified version of IPPR’s assessment debate over television. It is also possible to use (Coote & Lenaghan 1997: 11). videoconferencing and, in the future, virtual Four of the five IPPR juries were commis- reality systems, to offer electronically sioned by health authorities, the fifth by IPPR mediated face-to-face deliberation. Which is itself. The juries were set the task of address- better, physical or virtual interaction? Even ing important questions about policy and when people can get together in a neighbor- planning, for example “What can be done to hood, some may prefer to participate at an improve the quality of life for people with electronic distance. Further study is needed to severe and enduring mental illness, carers, and elucidate the pros and cons of these options. their neighbours?” (Coote & Lenaghan 1997: viii). The IPPR used a market research CITIZEN PANELS: UNITED KINGDOM organization to recruit and run the juries of 12 or 16 people. Two methods of recruitment In 1996 the Institute for Public Policy were used, each seeking to create a jury so that Research (IPPR) in the United Kingdom, as a it matched the population in terms of five pilot project, ran five citizen juries (called demographic variables: social class, age, “citizens’ juries” by the IPPR). Like those gender, ethnic background, and housing juries that preceded them, the IPPR series was tenure. “inspired by a critique of representative In Method 1, profiles were drawn up for democracy and a desire to rebuild public jurors so that the jury would statistically confidence in the political process” (Coote & match, as closely as possible, the population Lenaghan 1997: i). The Institute modeled its according to these five variables. For example, work on Crosby’s juries and Dienel’s planning one member might be expected to be a white cells. All five juries addressed health policy working class female, aged 20-30, in rental questions. IPPR chose the jury model because accommodation. Recruiters then would seek it uniquely combined the factors of informa- someone fitting this profile, going door-to- tion, time, scrutiny, deliberation, and door and approaching people in the street, at independence. IPPR assessed the extent to the same time searching for individuals fitting which other consultative methods matched the the profiles for the other jurors.

Table 1. Participation characteristics of various consultative methods

Model Information Time Scrutiny Deliberation Independence Opinion poll 0 0 0 0 0 Focus group 1 2 1 1 0 Referendum 2 2 1 1 3 Public meeting 2 2 1 1 1 Citizens forum 2 2 2 2 1 Deliberative poll 2 2 2 2 1 Citizen jury 3 3 3 3 2

Key: 0=no, 1=a little/sometimes, 2=some/usually, 3=a lot/a rule

Information: are the participants informed about the background to the policy question? Time: are the participants given time to consider the policy question before responding? Scrutiny: do the participants have the opportunity to ask their own questions about the subject matter and receive answers before responding to the policy question? Deliberation: do the participants have an opportunity to reflect on the information before them and discuss the matter with each other before responding? Independence: do the participants have any control over how their response is interpreted and communicated to others?

In Method 2, thousands of letters were sent to IPPR notes that both its methods of selec- local residents asking whether they would be tion involve an element of self selection and interested in being involved in a citizens jury, are not truly random. IPPR decided that “a but providing little information about the jury recruited to achieve some degree of process and none about the topic. Those who representativeness was less likely to incur responded and provided information requested criticism than a randomly selected jury that by about themselves were then put into relevant chance excluded one or more significant demographic profile groups. The actual jurors groups” (Coote & Lenaghan 1997: 73). It were selected by lot from these groups (Coote should be noted that self-selection is always & Lenaghan 1997: 71-72). involved unless jurors are conscripted, because The IPPR juries were brought together for prospective participants can accept or decline, four days with a team of two moderators; jury thus skewing the sample. members were paid up to £200 for their time. IPPR set the agenda, inviting and briefing the SHARED STRENGTHS AND witnesses. The jurors were given background WEAKNESSES information on the policy question, including written information and evidence from A stratified sample is necessary to compensate witnesses, up to four per day. Jurors were able for unequal volunteering rates and for chance to cross examine witnesses and discuss aspects deviations from desired demographic charac- of the topic in both small groups and plenary teristics though the importance of the latter is sessions. Since men often initially dominated arguable and may be issue-dependent. It could the plenary sessions, women and men were be said that none of the methods described separated in some of the small group discus- above involves unconstrained random selec- sions (Coote & Lenaghan 1997: 15). tion. One needs to distinguish between the The jurors compiled their conclusions in a various selection choices when one is report. The IPPR expected the commissioning discussing random selection. Firstly, obtaining body to do several things, including publiciz- a random sample of the entire population is ing the event and responding publicly to the probably not feasible because it would involve jury’s report within a set time. When the jury conscription. It is possible to come close to made definite recommendations, the commis- this ideal by randomly selecting, then sioning body was supposed to either act on following up personally to convince people to them or explain why not (Coote & Lenaghan participate (as occurred with the Australian 1997: 9). parent jury and the precinct committee). The organizers have drawn a number of Secondly, random selection from volunteers lessons from these citizen panels. Firstly, — the method used for the Danish consensus citizens are both able and willing to confront conferences — is feasible but can be criticized complex policy questions. Secondly, the for being unrepresentative. Thirdly, stratified citizen panel works better in choosing between random selection from volunteers — the well-defined options and in formulating method used by citizen panels — is feasible guidelines for decision makers than in dealing and commands respect for its representa- with abstract questions or producing detailed tiveness. plans. Thirdly, citizens, service users, experts, Some organizers have been reluctant to use and interest groups should be involved in the random selection because of its perceived consultation process; the consultation method cultural inappropriateness. Others feared chosen should be appropriate to the authority’s criticism about the non-representativeness that purpose. Fourthly, innovative models such as might occur by chance. In the examples above, citizen panels may need to follow an agreed organizers stratified their samples according to set of rules to ensure that they are credible and the sex of participants (to ensure gender reliable. balance) or the sector they would be from (a street, a school year, a health zone, welfare 60 Random selection in politics

recipients, and so on) to avoid the chance they can be susceptible to groupthink (a herd- occurrence of unrepresentative panels. Some like response) or to being swayed by a methods began with a profile and randomly charismatic individual. The methods outlined selected until the profile was matched. above have two components that guard against Random digit dialing could be seen to these possible responses. Firstly, they employ discriminate against those without a telephone trained facilitators to ensure that participants (the lowest socio-economic groupings), those are heard and that domination is minimized. unwilling to be forthcoming, and groups with Secondly, most methods lead towards a fewer telephones per person (for example consensus (unanimity), which also lessens the multiple occupancy). possibility of manipulation. And, of course, Arguments occur about the importance of the greatest strength of a small group is its randomness. Since selected participants are ability to interact constructively and achieve never going to be a perfect cross section of the worthwhile outcomes. entire population (just as a criminal jury is All of the methods mentioned above lack not), then is it not better to interfere in the official authority. The extent to which selection to ensure that as many interests as recommendations are acted upon is dependent possible are represented? Of course, a danger upon the sponsoring agency. The German emerges when one assumes that a woman, an planning cells (Chapter 5) are a good example indigenous person, or a person with a disabil- of positive influence, as was the Citizen ity is representative of the wider group. Survey Panel in Boulder. The Jefferson Center Feminists have wrestled with the shortcomings policy juries (Chapter 5) have had much less of seeing women as a homogeneous group, influence on policy makers. knowing for sure that the life of a single black This dependence on a higher authority to working mother in New York is very different implement recommendations points to the top- to that of a married white female with no down nature of political systems. Most of children in Kansas. these citizen participation initiatives were However, there is a commonsense promoted by government agencies or consul- response: elicit as many views as possible. For tation practitioners. The neighborhood policy example, when random selection is used to jury in Ballina is an exception to this but was select residents to consult about housing, still dependent upon a local council to take up ensure that both home owners and renters are its recommendations. Is the effect of such represented, that those who stay at home and consultation methods more on the participants those who work are represented, and that the than the policy makers? Participants clearly different needs of men and women with and enjoyed the experience and gained confidence without families are gauged as well. An from it. Assessment of the extent to which organizer would want a sample of the resident public involvement is a feel-good exercise or population that allowed as many views as an exercise in social action is clearly linked to possible to be canvassed. If, however, the the extent to which the entire community takes topic does not directly affect each participant such processes seriously. — for example, a question like child care — The agenda has most often been controlled then on what basis should selection occur? by organizers though some have had consider- Might it not then be appropriate to select from able in-built flexibility for altering that volunteers at random? agenda. Information was provided by the Random selection, particularly for small organizers and there were differences about groups, inevitably limits the chance of broad the way in which this information was community participation. The size militates compiled. Also, participants were often able to against this although some organizers have run call for more information. Groups were parallel groups to counteract this. A trade-off evidently well informed and this allowed for occurs between practicality and pluralism. quality deliberations. The sorts of information- Small groups have other potential weaknesses: sharing and information-seeking activities in Citizen participation with random selection: yesterday, today, and tomorrow 61

which participants were engaged varied. The therapy, or even manipulation (Arnstein activities included surveys, home or phone 1969). interviews, hearings, small group discussions, The evaluations completed by citizen jury videos, expert panels, written materials, field organizers indicate that randomly-selected trips, craft work, and visualizations. participants make “good” decisions. They are Randomly-selected groups offer a much likely to use common sense rather than legal better alternative when officials are forced to or other technicalities. Their recommendations negotiate an outcome which is agreeable to have proved to be useful for government opposing parties. Participants are far more officials. The average citizen is capable of objective and act as advocates for the larger handling quite complex information when it is society rather than a particular interest group. presented in an accessible, comprehensible Randomly-selected groups are obviously form. Their contributions can be creative, useful in settling contentious issues. However, fresh, and highly relevant and should be it would be unfortunate if this was the only valued. Payment is one way of both rewarding motivation for using them since their value in and valuing participants’ contribution as well timely, long-term planning seems to have been as a means of compensating them for time largely untapped. and/or wages foregone. Some of the methods The reality is that since government surveyed above involved payment to partici- officials cannot predict the recommendations pants, at least enough to cover their expenses; of randomly-selected groups, they are wary of others relied on volunteers. Payment is seldom using them. Any formalized citizen input is a a major motivation for attendance. It does threat to the power of government officials. allow participation by those otherwise unable Despite support for policy juries, for example to attend and shows that participants’ input is in the U.S. experience, they remain relatively valued. unused. There was considerable variation in the size Colin Brown (1997: 172) notes that of groups, from a cast of many hundreds in government officials value citizen input only deliberative polls, to a survey panel of 147, when decisions are “tricky” and suggests that down to the more common group of 12-20 in giving jurors “easy problems, where a number citizen juries. The time taken also varied, from of different options are feasible and accept- a year-long survey panel to a one-day or one- able,” would be much more worthwhile for the evening jury. Costs matched the scale — from decision-making authority. Brown notes that very small-scale, low-budget affairs to more there are two contrasting motivations behind expensive nation-wide events. Even large- initiatives for public involvement in decision scale citizen juries or planning cells are cost making. One is increasing citizen participation effective when compared with, say, a large- and the other is making better decisions. scale public survey that involves expensive Though complementary, they are distinctively design, implementation, analysis, and different, and go to the heart of any participa- reporting. tory process. Of course, helping authorities to Because randomly-selected groups are make better decisions is a significant goal relatively untried there is an absence as yet of because it leads to better policies and practices protocols. This is changing as more organiza- that in turn benefit citizens. However, having tions trial and evaluate these methods. The citizens make their own decisions is a very more experience that is documented, the easier different exercise. it will be for agencies to confidently employ There is an additional motivation: dampen- these strategies, and to be able to match their ing or co-opting citizen protest by channeling needs to a particular method. The processes it into safe avenues, thus legitimizing outlined above have been quite formal and decisions without seriously challenging power systematic and this remains a strength. relations. At its worst, citizen participation becomes tokenistic, or a form of placation, 62 Random selection in politics

THE FUTURE best features of Dienel’s planning cells with more traditional planning processes into a There is a danger in selecting a consultative three-step procedure, described below. This method as one would a consumer product — procedure is based on a belief that (a) “social off the shelf. It is better to think in terms of acceptance of any policy is closely linked with contextually appropriate strategies. Reading the perception of a procedure’s fairness in about the experiences of others alerts us to the making the decision” and (b) “the public has difficulty of transposing one situation on something valuable to contribute to policy another (Lane forthcoming). Context is crucial making” (Renn et al. 1993: 209). Random and being open to the uniqueness of each selection is a key component of two of the situation will mean the difference between three steps. effective and ineffective consultation and Though not neglecting the contribution of participation. Approaching each situation experts and regulators, Renn et al. seek to without preconceived ideas, wondering what draw on the competence and wisdom of the constitutes this particular community, will general public, particularly local knowledge achieve far more than deciding the method in held by community members. Their aim is to advance. take advantage of citizens’ full potential, Many of the methods that have been systematically combining citizen input with outlined above are top-down approaches to the contributions from experts and stakeholders. resolution of problems. That is, they are The three-step model draws on three kinds of methods designed to allow those with less knowledge, namely common sense knowledge power to have some influence on those with based on personal experience, expert more power, rather than to give decision knowledge based on technical skills, and making power directly to the currently less knowledge associated with interest groups and powerful. This should not negate the worthi- advocacy. ness of the methods but as such they may do Different groups are engaged in each of the little to change inequitable structures and three different stages or steps. In the first step, systems. Community members can choose to the objectives or goals are identified. This is exploit top-down approaches to achieve done by getting all relevant stakeholders — greater influence, while the decision-making namely, groups affected by the decision, or elite can use these approaches knowing that that perceive themselves to be affected — to they are bestowing a privilege on those who lay their values on the table and tell how they would otherwise be excluded. Government think different options should be judged (Renn officials have a choice: (1) to see the use of et al. 1993: 190). A value-tree analysis is used top-down approaches as a once-only opportu- to elicit these values and criteria. Individuals nity to make their own lives easier or to quell participating in this step can be selected by lot disquiet or (2) as a way to genuinely involve from relevant groups. the wider community in more participatory The second step involves either the research forms of decision making, ones which may team running the procedure or an external serve to transform current realities. With this team of experts. This team starts with the in mind, how might advocates of participation values and criteria for judging options derived operationalize the strengths outlined above in step 1 and transforms them into operational and begin to avoid some of the weaknesses? definitions and indicators. Then the group of What would an ideal (albeit top-down) stakeholders in step 1 reviews these results. consultative approach look like? Once they are approved, the agreed values and Ortwin Renn and others (including Peter criteria are used to evaluate the policy options. Dienel) have documented a process that has A heterogeneous group of experts judges the been used many times in West Germany as performance of each option, the aim being to well as the U.S. (Renn et al. 1993). The reach “an expert consensus via direct consultative process combines some of the confrontation” (Renn et al. 1993: 191). The Citizen participation with random selection: yesterday, today, and tomorrow 63

goal is to develop a profile for each option stakeholder witnesses are also involved. The telling how well it measures up according to sponsor has input but is not a central player. the agreed criteria. The research team co-ordinates the entire In the third step, each option profile is process, providing drafts and ensuring there is evaluated by a group of randomly selected approval of the joint value-tree, performance citizens, or sometimes by several such groups. profiles, and citizen report. This step employs processes similar to those Planning cells — the third step in a three-step used in planning cells (Chapter 5). Over three model — were initially used in urban planning to five days, the citizens learn about the policy when local governments were quite willing to options from lectures, hearings, panel give citizens an opportunity to contribute. discussions, and tours of relevant sites. They However, this receptiveness may not be evaluate the options, drawing on the profiles present on national issues with high stakes and developed in step 2. entrenched positions. Adding the first two Table 2 shows the involvement of different steps to create the three-step procedure may be types of “actors” — stakeholders, experts, more attractive and effective in these difficult citizens, sponsor, and research team — at each situations. The main opponents of turning step. Although one actor has the most decisions over to citizens are stakeholders important part in each step (bold face in Table who, especially in the U.S., expect to be 2), every group can be involved in each step. involved in decision making. In this context, For example, stakeholders are the main source one advantage of the three-step model is that of values in step 1, but others can contribute to stakeholders, experts, and citizens all have an establishing the value trees. Similarly, in step input (Renn et al. 1993: 199-200). In using the 2, performance profiles are primarily the procedures practiced by Renn et al., Carson responsibility of the group of experts, but (1998) found the combination of steps to be others can provide input. Finally, citizens are quite effective. the key group in step 3, but experts and

Table 2. Elements of Three-Step Participation Model (Renn et al. 1993: 192)

Actors Step 1 Step 2 Step 3 Concerns and criteria Assessment of Evaluation of options options Stakeholders (interest Establishing values Suggestions for Witnesses to citizen groups) and judgment experts panels criteria Experts Additions to concern Group delphi Participation as list (generation of collection of expert discussants or options) judgments presenters Citizens (randomly Additions and Transformation of Option evaluation selected) modifications of expert judgments and concern list recommendation Sponsor Input to concern list Incorporation of Witness to citizen (generation of institutional panels options) knowledge Research team Transformation of Verification of expert Compilation of concerns into judgments citizen report indicators Products Joint value tree Performance profile Priority of options for each option and policy recommendations 64 Random selection in politics

A number of steps seem to be missing, Ned Crosby, there has been a dramatic however, from the three-step model, though expansion in experiments with randomly aspects of these additional steps may have selected citizens being involved in making been unstated but assumed in Renn et al.’s judgments about specific issues following practice. For example, a significant first step exposure to information and a process of has been alluded to in relation to a number of discussion. Sortition may still be a neglected the methods already outlined: the early option, but it is fair to say that it is now being contractual period during which organizers given more attention than at any time since the and the final policy makers need to agree on rise of representative government more than their commitment to effect change. Whether two centuries ago. the consultative method is a community or So far we have discussed citizen participa- government initiative, it is essential that tion with random selection that has been organizers are able to inform participants promoted as a reform within present govern- about the possibility of any recommendations ment systems, though some advocates being enacted. Contracts need to be negotiated definitely have a long-term goal of direct and signed so that they are available to democracy. Proponents and experimenters participants. If participants know there is a low have different and sometimes multiple aims, probability of enactment occurring, then they including: are able to approach the task realistically, • influencing decision-making by govern- knowing that the advantages for them in ments; participating (education, skills-acquisition, • obtaining an assessment of what public curiosity, etc.) do not include the possibility of opinion would be in a society with widespread imminent institutional change. This step might deliberation on social issues; therefore be termed “contracting”. • balancing decision-making inputs from Part of this first step is also the necessity to experts and nonexperts; inform the wider public about the consultative • developing the interest and capacity of process. Random selection of small panels citizens to participate in making decisions that leaves little impression on the general public affect them; unless publicity occurs. Such publicity serves • reaching “better” decisions, especially a number of functions: to provide information ones reflecting the public interest rather than to the general public about the issue under special interests; discussion; to raise awareness about the • laying the basis for direct democracy random selection process; and to create some using sortition. interest in the possibility for selection. So “informing” is an important component of step To develop the last point in this list, we next 1 in this revised model. turn in Chapter 7 to some visions of a partici- patory political future integrally involving CONCLUSION sortition.

For several centuries prior to the 1900s, the use of random selection for political decision making was virtually nonexistent. The criminal jury provided the principal means for direct participation by randomly selected citizens. The rise of opinion polling in the 1900s provided a means for citizens’ views to be registered and fed into the political process. Then, in the last few decades of the century, beginning with the work of Peter Dienel and 7 Sortition futures

Direct citizen participation in decision making can be a valuable tool in planning, but what is valuable because decisions are more likely about actually running the society? to reflect people’s interests. Also, direct One solution is to reduce the size of deci- participation helps prevent domination by a sion-making communities (Dahl and Tufte ruling elite. Even when there are elections, the 1973; Kohr 1957; Sale 1980). Ancient Athens input of citizens into policy making may be so had no more than 60,000 citizens. If, today, limited that it’s almost like an elected dicta- communities of this size could be self-reliant torship. economically and politically, then they might As we described in Chapter 2, random well employ the methods used by the ancient selection played a key role in ancient Athenian Greeks, including a mixture of full meetings, democracy. Since then, the desire for popular random selection, and voting, with strong involvement in decision making has repeat- controls over all office bearers. edly surfaced in many guises, including mass While this might be attractive in some parts action against repressive regimes, the expan- of the world, it has a number of flaws. Few if sion of representative government, workers’ any communities in the industrialized world self-management, and citizen protests against are as self-reliant as ancient Greek city-states, government actions. However, the ancient so a self-contained decision-making system Athenians exceeded any society since in their doesn’t make as much sense. There is a much extensive use of random selection to run greater division of labor, with specialized society. In recent centuries, the criminal jury occupations from building to biomedicine. has been the only substantive decision-making Increased specialization also increases the entity drawing on random selection. requirements on decision makers. There are Beginning in the 1970s, political pioneers now many more issues requiring attention, Peter Dienel and Ned Crosby showed that often involving specialized knowledge randomly selected citizens are willing and able intermeshed with wider community concerns to participate in the sort of decision making and values. needed to run a society. In Chapters 5 and 6, If every citizen sits on a community panel we outlined the growing interest in and to decide on policy for every issue, everyone’s experimentation with citizen juries in several time would be taken up in decision making, countries. However, citizen juries remain, both with no time to do anything else. This can be in conception and practice, a supplement to the considered the fundamental problem of conventional political system based on elected participation: beyond a certain small size, not governments and standing bureaucracies. They everyone in a group can be involved in every show that much more citizen participation in decision. decision making is possible, but on their own There are several possible solutions to this go only a small distance towards structural fundamental problem. One is to leave decision change. making to a small group. In the case of For all its undoubted flaws, ancient Athens dictatorship, the rulers make up the small remains a model for the extensive use of group. In the case of bureaucracy — including random selection. But can the Athenian model large corporations — the top managers make be transplanted into contemporary societies in up the small group. In the case of representa- which there are vastly more people and where tive government, the top elected officials make new and complex issues, such as computer up the small group. The problem with these fraud and in vitro fertilization, need address- solutions is that participation is limited and ing? Citizen juries show that random selection power is concentrated. With representative 66 Random selection in politics

government, the problem of overload on Randomly selected legislators are less likely to elected officials still occurs, though it is be susceptible to special interests. If they are restricted to the top decision makers rather selected for one term of office only, then there than everyone. is limited opportunity for special interests to Another solution is to allow everyone to be cultivate relationships. involved in decision making, for example One objection is that some randomly through electronic referendums, as described selected legislators might be ignorant and in Chapter 3. The problem here is that many prejudiced. They might be venal and open to people will not be well informed — it is bribery. They might not take their jobs impossible to be well informed about every- seriously. But would this make them any thing — so decisions are made more on the worse than elected politicians? Politicians are basis of prejudice than careful consideration. well known to make compromises to be So, for any sizeable community, there endorsed and elected, and to spend a large seems to be a trade-off between two short- fraction of their time in cultivating favors, comings: fund raising, campaigning, and public • if everyone is allowed to participate in relations, so they can be elected and reelected. decision making, then many people will be Randomly selected legislators would have no uninformed, since there isn’t enough time for need for these activities. They could just get everyone to become knowledgeable about down to the job. After a short term of office, every issue; they would return to their communities and • if only some people make the decisions, have to live with their neighbors and col- then there is a concentration of power, which leagues, who would know how they had voted. may become self-perpetuating as decision Self-respect and peer pressure would be potent makers use their power to entrench themselves forces in reducing laziness and corruption. and their patrons. Another approach is to use random selec- Can random selection come to the rescue? tion to modify a process involving election or In this chapter we introduce ideas for alterna- appointment. For example, in selecting the US tives to electoral politics that go far beyond president, many of those with the greatest what many people see as possible or plausible influence — political party insiders, lobby today. Nevertheless, we think it is useful to groups, and candidates themselves — are far consider such utopian ideas, since sometimes more concerned about their own interests than they provide an effective guide to developing the general interest. Including a random and promoting alternatives in the here and element in the selection process would curb now. the role of self-interest. For example, having One idea is the citizen legislature. Rather the president selected randomly from the ten than representatives being elected, they could candidates who receive the highest number of be chosen randomly from citizens. Instead of a votes in an election would reduce the attrac- house of representatives, there could be a tion of putting big money behind a single “representative house,” namely a legislative candidate. Another possibility is that ten body made up of randomly selected citizens, regions of the country would each elect a rather like a large jury (Becker 1976: 467-470; candidate, with the president being chosen Becker et al. 1976: 183-185; Callenbach and randomly from the ten finalists. A U.S. Phillips 1985; Dahl 1970: 149-153; Mueller et Supreme Court justice might be selected al. 1972; Slaton 1992; Steele 1995). randomly from 50 nominees, one from each of A citizen legislature immediately solves the states, thereby reducing federal-level some of the problems of concentration of politicking and reorienting the court to states’ power. Prior to being picked, randomly rights. Indeed, there are innumerable ways to selected legislators cannot be endorsed, incorporate randomness in selection proce- promoted, or advertised. Therefore, the power dures, some quite complicated. The precise of political parties to set the agenda is reduced. Sortition futures 67

method used can be chosen to minimize the body to make decisions about every issue. role of political greed (Knag 1998). Could the same sort of approach apply to Still, selecting officials randomly doesn’t decisions that are currently seen as “political”? solve the problem of overload. When legisla- Let’s summarize where we’ve come to. tors, whether chosen by election or sortition, Decision making by a single body of officials have to deal with every issue, they do not have (elected or otherwise) has two major the time to become familiar with all the problems: abuse of power and overload of the arguments. The most likely result is sloppy decision makers. Using random selection decisions or dependence on support staff in reduces the potential for abuse of power. bureaucracies, thus leaving much of the power Dividing up the issues and assigning them to in the hands of special interests. One way to different groups reduces the overload. cut through this dilemma is to question the Combining these two methods gives a differ- assumption that a single decision-making body ent type of system. It is sufficiently different must deal with every issue. That’s the way from democracy that it’s useful to have a present-day legislatures and executives operate different name: demarchy. — and it leads to centralization of power. Why not break up the issues and have different DEMARCHY groups deal with each one? That’s the way much practical work is carried on. School The word “demarchy” is the creation of systems make decisions about curricula, philosopher John Burnheim, who developed construction firms make decisions about the idea in his pioneering book Is Democracy materials, and farmers make decisions about Possible? (Burnheim 1985; see also Burnheim sowing and harvesting their crops. Even 1986, 1990; for critiques see Lynch 1989; legislatures rely extensively on committees to Hirst 1986). Demarchy is such a change from deal with specific issues. the familiar representative government that it’s In everyday life, most people mainly focus worth spelling out the basic elements of both on issues with which they are directly systems for a systematic comparison. involved. They don’t rely on a single central

Representative government

Group Role in decision making How selected

elected politicians make decisions elected by citizens

government bureaucrats advise politicians; make senior bureaucrats appointed by administrative decisions politicians; others appointed by senior bureaucrats

political parties choose candidates voluntary membership

media influence opinion privately owned or government run

lobbyists and activists try to influence politicians, volunteers or paid by interest bureaucrats, and public groups opinion

citizens vote for politicians satisfy legal requirements (citizen- ship, voter registration, etc.) 68 Random selection in politics

Demarchy

Group Role in decision making How selected

members of functional make decisions random selection from volunteers groups

members of second- make decisions about the elected or randomly selected from order groups decision-making system members of functional groups

media influence opinion privately owned or run by community

lobbyists and activists try to influence members of volunteers or paid by interest functional groups and public groups opinion

citizens occasionally be a member of satisfy requirements set by second- a functional group order groups

In demarchy, there is no government “out considered to be just like a policy jury, except there,” administering everything for the that it meets on a semi-permanent basis. community. Rather, decision making is close Functional groups operate to serve the local to home, more on a scale akin to ancient community. This would probably be a popula- Athens. The key decision-making processes tion of thousands of people, rather like ancient occur in the “functional groups.” These are Athens. It might be 20,000, perhaps as low as groups of citizens, chosen randomly from 1000 or as high as 100,000 or more. Experi- volunteers, who make decisions in particular ence would provide guidance as to the areas — functions! Some possible functional appropriate scale for functional groups. So groups might cover: instead of a population of one million people • garbage collection; having a single high-level parliament or • child care; congress, perhaps with regional and local • music and theatre; government too, there would be a patchwork • industry policy; of communities, each with a set of functional • construction; groups. For a population of one million, there • roads; might be 50 or 100 such communities, each • public transport; with functional groups on all relevant topics. • energy; Let’s zoom in to a single functional group. • water; How big? Let’s say 12 people. Who are they? • research policy; Anyone could volunteer to be on any group. • health policy; The actual group members would be chosen • education. randomly from the volunteers. Are you So, instead of a single congress, parliament, interested in water policy? You might be or council dealing with all areas and running concerned about building more dams, about things through bureaucracies, there are many the levels of chlorination and fluoridation, functional groups, each dealing with a single about how water is paid for, about promoting function. Each functional group can be rainwater tanks and grey-water systems, about Sortition futures 69

water pollution, about dealing with drought, for learning without the risk of making ill- and a host of other issues. If so, you could informed decisions. volunteer for the group. Since water policy is The other safeguard is random selection. not just a local issue, there would need to be No one would be able to guarantee their place coordination with groups in neighboring on any committee. The number of people with communities. special interests in a topic such as water policy As we discussed in Chapters 5 and 6, will not be all that great, so before long others, members of citizen juries typically devote without an ax to grind, are bound to be themselves seriously and conscientiously to involved. their task, learn a lot, and come up with The process of random selection leads sensible sorts of findings. In the context of naturally to several results. People chosen for demarchy it is more accurate to refer to policy the groups are there not because they have juries rather than citizen juries since the special knowledge (though they might), nor groups make policy. Members’ expertise is because they have wide support in the initially quite limited, but even so policy juries community (though they might), but simply soon become, on a collective basis, quite because they were the lucky ones whose knowledgeable — certainly more so than the names came up in the draw. Since they are average citizen or even the average politician. there due to chance with no electoral mandate, A functional group would have the benefit of there is little justification for anything more this sort of special attention to the issue, plus than a limited term. This is similar to the two advantages. Firstly, the members would tenure of a criminal jury. Although jury have far more than a week or two to deal with members have more expertise after a case, the issues. Over months and years, they would they are turned out for a new jury. This is develop considerable insight into the issues. quite a contrast to electoral politics, in which Secondly, the members would be chosen from politicians often run for reelection. It has volunteers for that particular group. Those always been difficult to pass laws against with special interest and concern about a multiple terms of office simply because it is in particular issue would be much more likely to the interests of politicians to remain in power volunteer for the relevant group. If you are and the rationale of experience and popular interested in or concerned about water issues, mandate works effectively to keep this then it would make sense to volunteer for that situation unchanged. group, whereas if your main interests were in Sortition creates a political dynamic that is transport or child care, then it would be natural quite different from that created by elections. to volunteer for those groups. With elections, certain people desire to gain While a functional group might well have positions of power. Many are ambitious greater knowledge and experience with an people who seek power, fame, and wealth. To issue than a typical short-term policy jury, the build support within a major political party as corresponding danger is that there would be well as a wider voting public — the normal too great a bias due to development of vested road to electoral success — requires cultivat- interests. One way to overcome this is to have ing support from others for one’s own political strictly limited terms, perhaps one or two ambitions. The individual who does not seek years, with no possibility of reappointment. office but is pressured into standing for the The terms could be staggered, with a fraction good of the community is increasingly rare. of the members replaced at regular intervals, With sortition, there is no party or constitu- so that there is some degree of continuity. ency that needs to be mobilized to be chosen: Another useful measure would be to require it’s just a matter of chance. Furthermore, the new members to sit in on meetings as observ- limited terms of reference of any single ers for a few months before they can partici- functional group, and the limited term of pate in decisions. That would provide a period office, mean that there is little power, fame, or privilege to be gained even if one is selected. 70 Random selection in politics

Many people are unenthusiastic about serving to principles that might withstand scrutiny and on a criminal jury; likewise, many people provide guidance in new territory. might well be unenthusiastic about volunteer- In short, the politics of elections is one of ing for a functional group. Those who are ambition, special interests, and manipulation, ambitious would be better advised to seek with principles often subjugated to power- wealth or to exert influence by becoming an seeking. The politics of demarchy is one of expert in some field or a commentator in the education, mobilization, and issues of media. substance. However, there are some functional groups Let’s say that one particular interest group for which membership would be eagerly is highly effective in mobilizing supporters sought: those that deal with issues that excite and getting them to volunteer for a certain people’s passions. Topics such as abortion, functional group. Wouldn’t this mean that the guns, drugs, crime, and pornography would be group’s members would be atypical of the dealt with by some group or other. But how to community? Most of the volunteers — and, exert influence if selection is random? The therefore, most of those chosen randomly — obvious approach is to encourage as many might be men (or women), or managers, or people as possible to volunteer for the relevant from a particular ethnic group or religion, or group. It wouldn’t be effective, though, just to from the same neighborhood. Random get people to volunteer who agree with you on selection from volunteers might result in a the basis of a superficial familiarity with the very nonrandom group of people. issue. Group members would spend weeks and This was a serious enough problem for months studying evidence, hearing testimony policy juries. In demarchy, where the from experts and community members, and functional groups are the actual decision- discussing the issues before making a making bodies, and where canvassing could decision. To be an effective advocate for a occur to encourage people to put their names particular cause, a volunteer would need to be forward for the ballot, the problem is likely to an informed partisan, familiar with the be greater. arguments pro and con and able to deal with As before, one obvious solution is to draw a new evidence and understand new options. stratified sample, putting tight constraints on Thus, the best sort of volunteer to serve a the lottery to ensure desired characteristics of cause is someone who is knowledgeable, representativeness in the functional group. flexible, and yet committed to some general Consider a functional group for which it is principles. This means that any special interest thought highly important that half the group seeking to get more sympathetic people members be male and half female — perhaps onto a relevant group would need to engage in one dealing with family-related policies. a process of community education involving Suppose that 80 women and 20 men volun- those who are already sympathetic as well as teered for a ten-person group. To ensure equal those who are critical — since the latter might numbers of men and women, five members be selected too. The same would apply to a could be chosen randomly from the 80 women special interest group with an opposing view. and five from the 20 men. Alternatively, The result would be an intense process of suppose exact equality of numbers is not so promoting community education and important, but it is thought there should be at discussion. It would not be enough just to least one of each sex. Then the first nine sway people’s superficial opinions — the sort members could be chosen from the 100 of responses registered by opinion polls. This volunteers. If the nine include at least one man level of knowledge or commitment would be and one woman, then the last member would unlikely to withstand the intense scrutiny that be chosen from the remaining 91 volunteers. If would occur in a group. Instead, the aim the nine are all women, then the last member would need to be to develop a deep level of would be chosen from the 20 men volunteers. understanding of the issues and a commitment Similar methods could be used for other Sortition futures 71

grounds, such as on an ethnic, geographic, In demarchy, volunteering is for a more income, or religious basis. If some minimum specific task, namely a functional group on a number of members is desired from any topic such as power supply or the arts. Since particular group, this is easily arranged. In most volunteers will have a special interest in many ways, this is not greatly different from the topic, this is likely to improve efficiency representative government, in which there are but at the possible expense of wider participa- often seats designated for particular regions tion on any particular function. Following the (such as two senators from each state in the model of ancient Athens, it would be appro- U.S. Senate) or, sometimes, ethnic groups priate to institutionalize scrutiny of group (such as in Fiji). members before and after their terms. The reason for this sort of quota system is Objections to particular individuals, on that some types of people are more likely to specified grounds, could be considered by a volunteer than others. In many cases it is men, meeting of citizens if a sufficient number of the well educated, those with higher incomes, them petitioned for recall. At the end of a and those from dominant ethnic and religious member’s term, there could be a public groups. There is an alternative to volunteering hearing on their performance. There would and quotas, of course: including everyone in need to be a balance between, on the one hand, the random selection process, like the way sufficient scrutiny to promote and ensure juries are selected now. This would certainly adequate performance of duties and, on the produce a fair distribution of members from other hand, sufficient tolerance and support for different sectors of the community. The group members so that citizens are not disadvantage is that many people wouldn’t discouraged from volunteering. want to sit on a functional group, especially on a topic which they didn’t care much about — SECOND-ORDER GROUPS just as many people today try to avoid jury duty. Perhaps in highly participative, inte- So far we’ve talked about demarchy as a grated communities, with a well developed system of randomly selected groups that make communal sense of obligation, it would work decisions about particular issues such as to include everyone in the selection process industry or education. But there are quite a for every group. But without such a level of number of other sorts of decisions about the community solidarity, random selection from way the system is organized. For example: volunteers will result in groups whose • How is the number of members in a group members are far more committed to doing a decided? good job. Then, because volunteers may be an • How is it decided which particular groups atypical sample of community members, should exist? For example, should there be a specifications can be put on the composition single group dealing with industry, or separate of the group, with appropriate sortition groups dealing with agriculture, heavy procedures to meet these specifications. industry, and light industry? Should there be a In ancient Athens, extensive participation separate group dealing with a key local was fostered through sortition at the possible industry? cost of efficiency. Citizens of Athens • How are quotas — such as minimum volunteered for service and were scrutinized numbers of men or women — for groups before and especially after their term of office. decided? By putting tight controls on the powers of In representative systems, these sorts of randomly selected officials, and drawing equal issues are usually decided by the government numbers from each of the 10 tribes (the key itself, such as when a parliament decides to form of stratification), Athenian democracy change the voting system or the number of its was able to use sortition in a way never members. Sometimes decisions are made by equaled since. the courts or by statutory bodies such as an 72 Random selection in politics

electoral commission that decides boundaries demarchy that still need to be worked out. It is for electorates. possible to propose ways to deal with these To deal with these sorts of issues in aspects, with the understanding that many demarchy, John Burnheim came up with the things need to be worked out through trial and idea of having “second-order groups,” which error, using an experimental approach. are groups that make decisions about how the Blueprints can be useful to stimulate thinking demarchic system operates, dealing with but may not be accurate guides to what will questions such as those above. work. Here we will make a few comments on Burnheim suggested that members of various issues. second-order groups should be chosen from • Should membership on a functional group those who have served on functional groups be equivalent to a full-time job, a half-time — the first-order groups. How? Since the job, or just an occasional activity? It might be second-order groups require a lot of wisdom to more attractive to community members if it keep the system running smoothly, Burnheim was like a fractional appointment, a certain proposed that members of first-order groups number of days per week or weeks per year. should give confidential assessments on their • How many groups should there be? This peers’ suitability for higher-order tasks. Those partly depends on the need for decisions and who get the best ratings would go into a pool coordination. Consider, for example, a from which the actual members of second- community of 10,000 people. Suppose one order groups would be selected randomly. quarter of the people are interested in serving However, one argument against this is that on a functional group for, on average, one out selection by colleagues — even with the of 10 years (perhaps as a fractional appoint- element of sortition — would inhibit the free ment). Then there would be 250 people flow of discussion, since those seeking to get available at any given time to serve on on second-order groups might seek to cultivate functional groups, enough for 25 groups with favor with others. An alternative is direct 10 members each. In this picture, lots of random selection from members of functional people play a major role in community groups. There’s no easy way to judge between decision-making but at any given time only these alternatives. The obvious solution is to one out of 40 people is involved — not a large try out different options and see which one burden. Indeed, by having more groups, the works better. level of participation could be increased by a factor of five or more without an intolerable QUESTIONS ABOUT DEMARCHY burden. If there get to be too many groups or if not enough people volunteer to make groups The basic structure of demarchy is fairly clear. viable, then the second-order groups could In a community of thousands or tens of step in to eliminate or combine some groups. thousands of people, there are functional • Would the groups need support staff? It groups covering a range of important areas, might be useful for each group to have one or each one making policy decisions. Members two people who would call meetings, collect of the groups are chosen randomly from information, arrange for visiting experts, and volunteers. To negotiate the structure itself, so forth. These facilitators would need to be there are second-order groups, whose neutral with regard to the issues dealt with by members are chosen from those with experi- the group, just like the role played by ence on the functional groups. Just as impor- Jefferson Center staff in running policy juries. tant as the structure is what does not exist. A concern is that the facilitator or steering There is no central decision-making body, committee has a disproportionate amount of neither locally, regionally, nor nationally. power. One recent British consensus confer- There are no government bureaucracies to ence demonstrated that a lay panel feels execute the decisions of the functional groups. sufficiently empowered to work alone if its This means that there are a lot of aspects of members believe, rightly or wrongly, that the Sortition futures 73

group’s facilitator is biased (Simon Joss, to say that there are many options and many personal communication, 1997). Contrary to questions to be answered. the usual way of doing things, this group • How would coordination occur between simply prepared its own report, thereby groups? Groups might keep in contact with rejecting the external support provided to it. groups dealing with similar issues. They might Steering committees for processes such as hold joint meetings. They might refer disputes consensus conferences or citizen panels, in the to second-order groups. There are many experience of the authors, are kept honest by unanswered questions here. the selection of a group with a diverse and One important question is how to avoid balanced range of views. Human participants groups attending to their own little patch of are inevitably biased but processes can be put concern and not taking into account the wider in place which allow for exposure and context. For example, a group looking at water acceptance of these biases. policy would need to look beyond the local • What about people not on groups? Could geographical boundaries in order to consider they participate in any way? If they have wider ecological issues, future generations, relevant expertise or practical involvement social impacts, and many other factors. with the issues, they could appear before Coordination between groups is essential, but appropriate groups as experts or partisans. more is required to deal with this possible They could make written submissions to problem. groups, write letters to newspapers or on email • How would groups enforce their discussion groups, do door-knocks, produce decisions? The answer is that they wouldn’t leaflets, circulate petitions, hold rallies, go on and couldn’t! Here is where the deeply radical strike, join boycotts, and perform civil nature of demarchy becomes most apparent: disobedience. All the methods of nonviolent there is no authority that can exercise coercive public participation would be available, just as power. Under a representative government, they are supposed to be available in represen- decisions are normally implemented by tative systems. Such methods of participation bureaucrats. If there is significant resistance, it might well be even more influential in can be overcome by use of courts and, if demarchy, since members of functional groups necessary, police and military power. States would have no mandate to follow some are, by their nature, systems relying ultimately particular policy or party line. on the use of force to maintain power, though • How would the economic system operate? in practice they seek to foster popular support It might be a market-based system with small or acquiescence, which usually works far enterprises or it might be a more cooperative better. The “police power” of the state is system with collective provision of goods and apparent in the use of police (and sometimes services and worker-managed operations, the military) to arrest and imprison those who among other possibilities. Large multinational challenge the system in a fundamental fashion. corporations, in contrast, would not mesh well In capitalist systems this power is typically with demarchy. If big, remote government is used against challenges to private property, replaced by functional groups, then it would such as squatting in empty building or strikes be incongruous to still have giant corporate in essential industries. It is also used exten- bureaucracies, which are less accountable to sively in wartime to impose conscription and workers than governments are to citizens. repress dissent. Perhaps large corporations should themselves In demarchy, there is no state and no be run, in part, on demarchic principles! “police power” (though there might be police Certainly corporate policy, planning, to deal with ordinary criminal behavior). How evaluation, and dispute resolution could be then can decisions of functional groups have done this way. any impact? The answer is that they have Burnheim has ideas for using functional impact firstly because of groups’ credibility as groups to regulate the money supply. Suffice it representatives of the community who have 74 Random selection in politics

studied the issue in depth and secondly to the refuse to obey common standards embodied in extent that their decisions seem sensible and laws. Actually, though, a great part of social compelling. This is similar to the current role life proceeds on the basis of cooperation or of the criminal jury. Juries have credibility agreement about principles of behavior. because they are made up of community Societies are sustained largely by acquies- members. In most cases jury decisions are cence or support from relevant groups rather unquestioned. Occasionally a decision gains than the threat or exercise of brute force publicity because it seems unfair. Juries (Edelman 1971; Gramsci 1971; Sharp 1973). maintain credibility to the extent that they are Although law is often seen as the basis for seen as making sensible decisions. order, in practice community members may Functional groups would have several maintain order without or despite the law advantages over juries. They would not be (Ellickson 1991). This reliance on social hobbled by all the restrictions imposed on cooperation is quite obvious in demarchy. juries. For example, they would be able to For example, non-binding arbitration has seek out information themselves. They would been used to settle disagreements. Arbitrators be able to spell out their reasons for making a are agreed to by the contending parties, who decision, at some length. The more controver- are expected to adhere to the decision of the sial the topic, the greater the need for a careful arbitrators. There are no laws enforcing this explanation of the rationale for any decision. sort of arbitration (which has no legal or other Even when governments exercise coercive formal guarantees); however, a party that goes power to enforce their decisions, they cannot against an arbitrator’s decision would lose succeed in the face of widespread popular credibility with others, and probably would resistance. Prohibition has never worked when lose business and support. With a reputation lots of people want to take drugs. Likewise, for not holding to promises, such a party government controls over guns do not work would have a difficult time finding anyone when lots of people want their own guns. else to enter into an agreement (Watner 1997). Governments cannot just make any decision • What about defense against aggression? If they like and make it stick: they have to be demarchy has no state and no government sensitive to popular sentiment, otherwise they bureaucracies, that means it has no military may lose power due to citizen resentment and forces. One solution to this problem is discontent. nonviolent community resistance to aggression These same processes would operate under using methods such as rallies, strikes, demarchy, but much more strongly. A rogue boycotts, and sit-ins, with social and techno- functional group that made a decision that logical systems designed to support such a convinced nobody would have little influence resistance. This is called social defense, — people could just ignore the decision and nonviolent defense, or civilian-based defense carry on the way they preferred. If a functional (Boserup and Mack 1974; Burrowes 1996; group or other organization took action that Martin 1993; Randle 1994; Sharp and Jenkins adversely affected others — such as causing 1990). Although no society has ever organized pollution above widely agreed levels — then itself for social defense, there are quite a various forms of nonviolent action could be number of historical examples that suggest the used to address the problem, such as alerting power of nonviolent action, such as the the population, convincing others to refuse to Czechoslovakian resistance to the Soviet-led supply goods or services, instituting a boycott, invasion in 1968, the toppling of the Marcos or organizing rallies or sit-ins. To be effective, dictatorship in the Philippines in 1986, and the functional groups would need to rely on their collapse of Eastern European regimes in 1989. credibility as nonpartisan citizens and their Social defense relies on the same sort of good sense as to what is a workable decision. community participation and solidarity that is It is a common belief that society requires fostered in demarchy. It essentially means coercive governments to control those who design of social systems and social mobiliza- Sortition futures 75

tion to defend those things in society that people think are worthwhile.

CONCLUSION

Demarchy is a vision of a society that is participatory, eliminating the hierarchical structures of the state and bureaucracy, while overcoming the problem in direct democracy that people don’t have enough time or expertise to make decisions about every issue. The best evidence for the potential viability of demarchy is the experience with policy juries. Nothing like demarchy exists today. That is not a reason to reject it, any more than the absence of democratic systems in the year 1500 would have been a reason for rejecting democracy. Demarchy is a possible model for promoting participatory politics in a complex society. Because demarchy is only an outline of an alternative, much examination and experimentation is needed to see how it might be developed and improved. Just as the system of representative government requires a lot of fine tuning to ensure fair elections, controls over the executive and so forth, so demarchy will require many ad hoc adaptations to become a viable form of political life. Sortition is central to demarchy, ensuring that special interest groups cannot gain an entrenched hold over decision making as well as opening up participation without the carrot of ambition or the stigma of defeat that is so characteristic of electoral politics. 8 Strategies

Random selection in politics seems to have a Emery in 1976 proposed a system of lot of potential. It fosters participation, participative democracy based on juries undermines the entrenched position of operating in networks at conventional levels politicians and bureaucrats and is widely (local, town, regional, national), with people at perceived to be fair. But being a good idea higher levels selected randomly from those at isn’t enough. To be introduced, there needs to lower levels (Emery 1989c). be a strategy for promoting random selection. Although choice of strategy depends on the To develop a strategy, it is valuable to goal, there is actually a fair bit of common know the goal. Several different goals can be direction in these different goals. For example, discerned in work by proponents and users of promotion of demarchy is likely to both aid random selection in politics. and be aided by operation of policy juries. Reform of the electoral process. The Hence, treating various forms of random Jefferson Center’s electoral juries, in which selection in politics together may be satisfac- randomly selected groups of citizens study tory for a preliminary general discussion of party policies and question politicians in order strategy. to recommend particular candidates, are one Since policy juries, demarchy, and most example. Building on this experience with other uses of randomness in politics are not electoral juries, Ned Crosby (personal well known, there is not a lot of experience in communication, 21 January 1999) is preparing promoting them. Therefore we cannot give a to run “citizens election forums,” in which comprehensive assessment of strategies. citizen juries would evaluate and rank Nevertheless, there are some points worth candidates, with ratings widely distributed to making. We begin by outlining the most potential voters. He hopes to have this reform important sources of opposition to and support adopted through the initiative and referendum for random selection, then look at possible process. opportunities for promoting it, and finally Reform of policy making. Examples include discuss some general principles. planning cells, citizen juries, and Danish consensus conferences which are now using SOURCES OF OPPOSITION random selection. Members of consensus panels also have a greater control over the Sortition is a definite threat to all those who agenda, making this a more potent reform. gain power through some other mechanism. Reform of direct democracy methods. As Politicians and political parties rely on the discussed in Chapter 3, randomness can be electoral system for their legitimation and used to improve the operation of initiative and power. Their skills and organization are referendum, voting in face-to-face meetings, geared to winning support from voters, for and consensus decision making. example through public relations, campaign- Introduction of sortition-based alternatives ing, policy making, deals with interest groups, to representative government. Various models and ties to government bureaucracies. Holding of society involving sortition are possible. office allows the exercise of political patron- Examples include ancient Athenian democracy age, with benefits for favored individuals, (Chapter 2), citizen legislature (Chapter 7), organizations, industries, and sectors of and demarchy (Chapter 7). Ned Crosby in society. Holding office also gives politicians 1980 postulated a system built around citizen considerable visibility and status. It is for juries controlling government bureaucracies, precisely this reason that certain personality but was unable to get his book published. Fred types are drawn to electoral politics, especially Strategies 77

those who are ambitious, competitive, and Judges can be opponents of juries usurping good at making deals and portraying a positive their power. Mueller (1997) notes that antitrust image. legislation in dozens of countries has been left It is hard to imagine many politicians unenforced due to judges interpreting the laws willingly turning over all the advantages they in a way favorable to monopolies. Mueller have gained through electoral politics in favor argues that citizen juries would be less of a process of random selection in which they susceptible to the propaganda from the would have no more chance than anyone else. wealthy and should be in charge of antitrust No, the most likely responses are total lack of cases. interest, failure to even understand the alter- Lobbyists on behalf of groups such as native, and active antagonism. The lack of doctors, the telecommunications industry, or interest is apparent in the failure of govern- farmers have a stake in the electoral system ments to promote experiments in citizen because they have privileged access to participation that might replace some of their particular politicians and bureaucrats. By own functions. There have been many threatening to use their economic or voter successful planning cells and citizen juries power, or by making donations, they have an over a period of decades, but politicians have inside track to gaining advantages. Many not come knocking on Peter Dienel’s and Ned interest groups make donations to all major Crosby’s doors asking how to implement them political parties and keep on good terms with on a wider scale. As for demarchy, it is so key bureaucrats, many of whom remain alien to the consciousness of most politicians whatever happens at election time. The that they would be unable to grasp it easily, as electoral system provides a degree of predict- we’ve noticed in a few conversations. After ability for applying pressure. In contrast, it is all, it involves getting rid of governments as much more unreliable to try to influence a we know them. committee of citizens selected by lot. Even if There’s nothing new in this. Politicians and it was possible to make headway, the whole political parties are also resistant to other process would have to start again when new measures that might undermine their power, members are chosen. measures which could make the present Establishment experts also have a stake in representative system more accountable or the system. For example, the practice of fair. Proposals they have resisted include medicine is licensed by the state; health limited terms of office, setting of salaries by insurance systems provide payment only for an independent tribunal, the right of recall, certain categories of registered practitioners. limits on donations to parties, removal of party The connection between the medical estab- affiliation from ballot papers, and rotation of lishment and the state is the result of a long names on ballot papers. process of political mobilization by doctors. Politicians are likely to be among the most Introducing random selection might well vehement opponents of random selection, but destabilize this connection; establishment there are others who are also likely to be medical experts would have to make their case resistant, including government bureaucrats, to citizens on its merits rather than relying on judges, lobby groups, and establishment government endorsement. The same sort of experts. Government bureaucracies are thing is involved with lawyers, engineers, insulated from citizen input, which has to psychologists, and many other experts who are operate through voting and pressure group licensed by the state and whose advice is open politics. Sortition would open up a different to challenge. sort of accountability that would be It’s worth mentioning one additional source unwelcome especially to top bureaucrats who of resistance to random selection: many are intimately involved in formulating as well prominent figures in social movements and as implementing policy. dissident political groupings. In terms of their own principles, many of these individuals are 78 Random selection in politics

committed to promoting increased citizen system is okay and that what needs to be fixed participation. However, they occupy positions are errant individuals or bad policies. The in which they have status and power within an electoral system as a method of governance is organization and perhaps wider visibility as almost never questioned in any fundamental spokespeople for a cause. To promote random way (Ginsberg 1982). Even to change the selection might undermine their own status system of voting, for example from single- and power. This may lead to a lack of enthusi- member electorates to mixed-member asm for random selection and support for the proportional representation — as occurred in electoral system. Their disapproval of the New Zealand in October 1996 — is rare. status quo may be more about who is in charge Deeper changes are not on the agenda so far as and what policies are implemented than about the media are concerned. the decision-making system itself. State- Finally, the present system is a part of most oriented socialists are likely to be especially people’s lives. Watching and discussing the antagonistic to sortition, given their commit- activities of prominent politicians is a popular ment to a “party line” and belief in central spectator sport. Voting is a ritual for many planning. people. Others join the political process by We’ve commented that there is an obvious writing letters to politicians, participating in link between opposition to random selection organizations that apply pressure to and vested interests in the current system, politicians, or joining rallies and other protest whether it is politicians, judges, bureaucrats, actions aimed at influencing political experts, or leaders of dissident groups. decisions. Generally, “political” activity is However, most opponents would explain their assumed to be concerned with what govern- opposition using rational arguments. Individu- ments do. The lack of participation in decision als are not consciously biased. Rather, they making at work and in local communities is think in certain ways that often happen to ignored; it is just assumed that participation is reflect their situation in life. something only for the “political” sphere. Promoting random selection will be diffi- It’s useful to think of sources of resistance cult because of opposition from vested to sortition, because it can be a waste of time interests, but there is a deeper obstacle as well: trying to convert politicians and others with a the entrenchment of the current system in vested interest in standard methods. Also, it’s people’s minds and behavior. wise to be prepared for attacks. At school, children are taught that “democracy” — namely representative SOURCES OF SUPPORT government — is the best political system. There is no serious discussion of possible While opponents of random selection are often alternatives that might be more “democratic,” linked to vested interests and familiar ways of namely more participatory. behaving, it is harder to find an explanation The mass media also foster the idea that for why some people support random “democracy” is the ultimate political system. selection. Here are some possibilities. There is an intense focus on the political Firstly, many social activists are committed process, especially the personalities of to participatory group dynamics, for example political leaders, their jockeying for political the use of consensus in affinity groups, often advantage, and their struggles over policies. because they themselves are seeking the Quite a lot of coverage is a direct result of satisfaction that comes from direct involve- public relations, especially by governments. ment in issues. Many such activists reject The media often show the seamy side of electoral politics as a sham, having seen the politics, including the peccadilloes of failure of the political system to deal with politicians and factional in-fighting, and environmental problems, male domination, or sometimes expose pay-offs and cover-ups. The some other problem they have experienced underlying assumption, though, is that the personally. Some of these activists are actively Strategies 79

searching for participatory alternatives to to increase the role of discussion based on electoral politics and are open to the idea of more than self-interest and power-trading random selection. (Burnheim 1995). Thus, promoting random Secondly, there is a disparate group of selection through considered argument is individuals who are promoters of participation. nicely compatible with the goal of using This includes some scholars, community random selection to foster participative and workers, and activists who overlap with the deliberative decision making. In other words, social activists mentioned above — and a few the means (considered argument) is compati- elected representatives. They have developed a ble with the end (sensible decision making), commitment to participation and thus may be unlike familiar examples such as defending the open to random selection as one option. The peace through armed force. pioneers of planning cells and citizen juries fit Arguments on their own are seldom enough in this category. to win the day. Ideas are never autonomous of Thirdly, there are various people who are in social location and context. A sure sign that a search of a decision-making procedure that is person has a deep-seated emotional resistance widely seen to be fair. This includes some to an idea is when, after giving good answers government workers, corporate executives, to their objections, they keep responding with and community activists who are seeking a new objections and changing the terms of the way forward on difficult decisions that have discussion. the potential to cause serious rifts. Random We have already outlined some of the selection may be supported because it solves a arguments for and against direct democracy particular problem, though it might not be (Chapter 3) and random selection (Chapters 2, supported aside from this. 5, 6, and 7). Rather than rehearse these again, It is early days yet to know the basis of we will just revisit a couple of crucial support for random selection. If citizen juries concerns about random selection. ever become established, even in a restricted One objection is that random selection fashion, then sources of support will become means that the most knowledgeable and more obvious. This is apparent in the case of experienced people may be omitted from criminal juries, which are backed by a wide decision-making bodies, whereas the ignorant, range of legal scholars, lawyers, and citizens, prejudiced, and uninterested may be chosen. as well as providing direct experience for This concern can be voiced by those who jurors themselves. Until similar experience in support establishment experts as well as those citizen juries becomes widespread, support is who have developed expertise that challenges likely to be limited and precarious, confronted the establishment. This objection reflects a as it is by vested interests and personal deep-seated tension between expertise and experience of electoral politics. participation. In technocracy — rule by experts — participation is by experts only, ARGUMENTS since they are presumed to know best for everyone. In various forms of democracy, Part of the struggle for a new system such as expertise is subordinated to participation, demarchy is developing persuasive reasons, including participation by non-experts. arguments, and examples. Some opponents One response is that experts are not neces- may be motivated by vested interests, but they sarily the best people to make decisions that are bound to justify their positions by rational involve broader issues. Expertise is often very arguments, or at least ones that sound rational. narrow in scope. Does expertise in nuclear Developing the case for demarchy and engineering by itself provide a suitable formulating responses to objections is of qualification to set energy policy in everyone’s crucial importance. interests? Does expertise in econometric One of the main advantages of sortition is modeling by itself provide a suitable qualifi- to reduce the influence of vested interests and cation to decide on how a society’s resources 80 Random selection in politics

should be allocated? After all, few politicians the combination of voting and representatives have relevant expertise in nuclear engineering for decision making. Is there some analogue or econometric modeling. for random selection? Even when experts are not on decision- One way to start is to use random processes making bodies, they can still have a significant more often in small groups when fairness or influence on decisions. They can testify to equalizing power is a key consideration. For policy juries, give public lectures, write example, in groups using consensus, it is articles, and talk to whomever they wish, standard procedure for the task of facilitation including the media. In this model, experts are to be rotated around the group. This could effective by being persuasive rather than easily be replaced by a lottery (with no one through their formal position. This is the idea chosen twice until everyone has been chosen that experts should be “on tap but not on top.” once), thereby avoiding any subtle bias by Lots of the arguments against sortition boil which certain facilitators are chosen for certain down to resistance to participation and distrust types of issues. of ordinary people, whether the arguments are The idea of functional groups, the second couched in terms of expertise, merit, experi- main element in demarchy (the other being ence, or mandate and whether the method of random selection), can also be approximated selection is appointment, competition, or in small groups. Consider 20 or 30 people, election. committed to consensus, who are faced with At the opposite end of the spectrum is a making decisions about five or 10 topics in an different objection: that those who are afternoon. If every topic is dealt with by the interested, educated, and confident are much full group, the process of reaching consensus more likely to volunteer for a position chosen can sometimes be agonizingly slow. An randomly, leaving out those who are most alternative is to break into small groups, with disadvantaged. The alternative, requiring each group dealing with one topic (or more, if everyone to participate in a lottery, introduces there’s time), with the task of preparing a compulsion with all its nasty side-effects. The proposal for the full group. If there are no most obvious response to this problem is to volunteers to deal with a particular topic, then stratify the sample sufficiently — by income, perhaps it is not so important! We have seen employment status, education, or whatever is this process work quite effectively, generating required — so that people from any specifi- greater commitment to outcomes, speeding up cally defined disadvantaged group are decision making, and reducing aggravation. selected. In addition to this are methods for This is not the same as the functional groups encouraging participation, such as adequate in demarchy, but there are similarities. pay and life support for participants, education Random selection can be added in. For for participation, a supportive community example, if nearly everyone wants to deal with climate, and tolerance. These and other a particular topic, some can be chosen by lot foundations for participation are much the and the others can choose another topic. By same whether they are achieved through self- introducing techniques such as these, people managing groups, voting, or demarchy. can get a personal feel of the elements of demarchy, making it much easier to grasp the POSSIBLE OPPORTUNITIES wider picture. Only a few people are attracted to demarchy solely by reading theoretical To move towards a wider use of sortition, it is accounts such as John Burnheim’s book Is easier to start small. Policy juries are much Democracy Possible? (Burnheim 1985). easier to introduce than a full implementation Even to create small-scale experiences of of demarchy in a community. Indeed, one of random selection and functional groups can be the big problems facing demarchy is the scale quite difficult. Just asking a meeting to break of the operation. Voting can be used in a small up into small groups can generate resistance, group; representative government is simply especially from those who usually dominate Strategies 81

the discussion. The official heads of groups supplies, there have been hundreds of local have the best opportunity to experiment with referendums, which provide a convenient way different methods, though often the least for governments to off-load responsibility for incentive to do so. This includes teachers of making a decision that may alienate one classes, who can use random selection and section of the population (Crain et al. 1969). A small groups as innovative teaching methods, policy jury could be seen as another suitable and chairs of formal committees, who are in alternative for such issues. the best position to argue for experiments. It A large social movement committed to will be quite a while, though, before the chair participatory politics could introduce random of the legislature is able to introduce random selection and/or functional groups as part of its selection for membership of powerful own process of involving members and committees! developing sound policies. Portions of the As we described in Chapters 5 and 6, second-wave feminist movement, nonviolent various people have tried out policy juries in a action movement, and environmental range of circumstances. What can we learn movement were pioneers in promoting partici- from this experience? Firstly, even the best patory mechanisms including affinity groups designed and well run juries can easily be and consensus. It would simply be another ignored by the political mainstream. The stage for a movement to start “living the Wuppertal and Jefferson Center initiatives alternative” of sortition. continued for decades with little apparent While all these are possible ways for the impact beyond their participants and sympa- spreading of random selection in politics, they thetic observers. Secondly, some level of will depend crucially on two things. The first resources is needed to run the juries, which is a core of committed individuals who will can be provided through a university, take on the task of promoting the alternative. independently funded center, or local govern- The second is a change in the social climate so ment position. Thirdly, participants usually that it is more receptive to this particular think highly of the experience but seldom alternative. These two things go hand in hand: become ongoing advocates of the process. the committed individuals help to change the There are several conclusions one can draw climate and an altered climate will make the from this. One is that the time is not (yet) ripe efforts of individuals easier. Ultimately, for a wider uptake of random selection. random selection needs to become part of a Another is that efforts should be directed at social movement. more promising avenues, where random selection will be welcomed because it is seen PROMOTING RANDOM SELECTION to be fair. One possibility is organizations that are in a Although the wider use of random selection in deep crisis of confidence, perhaps due to politics is likely to depend on a change in scandals or poor performance (Burnheim political climate and favorable circumstances, 1985). This might be a school or hospital, for there are things that can be done in the example, where there is an assumption of meantime. Here we outline some things that serving the public interest. If the crisis causes we think are important if the full potential of serious divisions in the community, then random selection in politics is to be achieved. politicians may fear to intervene since they might be tainted no matter what they do. Make the Idea Credible Setting up a policy jury could be seen as a safer option. There’s not much chance of random selection Another possibility is that a progressive being taken up unless people know about it local government might set up policy juries as and consider it to be a feasible alternative. a way of dealing with contentious issues. In Plenty of opportunities for introducing it have the debate over fluoridation of public water come and gone because no one knew about it. 82 Random selection in politics

Hence one of the most important things is to closing off options. Opportunities may arise in raise the idea in all sorts of forums. This unexpected circumstances. It’s worth being includes talking with friends or at meetings, experimental. It might be that a long shot turns proposing random methods when decision- out to be a winner. making procedures are under review, sending There’s a tension here: new ideas and letters to newspapers, commenting on radio, experimentation are about opening up options preparing manuals, establishing web sites, and and, in some cases, proceeding in the face of producing videos, among others. criticisms, whereas critique tends to look at Some ideas have strong backing from shortcomings and to establish preferences vested interests, which can fund journals, think between options. If both experimentation and tanks, and front groups. Random selection has critique are valuable, then this tension is no such backing (at the moment!), so the task inevitable. Living with it is part of the of promotion falls on committed individuals challenge of promoting random selection. and groups. (Does it have similarities to living with the uncertainty of the outcome of a lottery?) Be Critical Maintain High Standards Fortunately, there is no official line on random selection in politics. Instead, various ideas are In running policy juries and other exercises being tossed around, from electronic polling to involving random selection, it is vital to do demarchy, without any central authority or them well. Poorly designed and run experi- guru to keep advocates on the straight and ments can be used to discredit the whole narrow. Key figures like John Burnheim, Ned approach. The work at Wuppertal and the Crosby, Peter Dienel, and Fred Emery each Jefferson Center has set an extremely high have their own preferred directions, to be sure, standard. Projects are well designed, planned, but none has tried to impose a “random and executed. This ensures a favorable orthodoxy.” This is not just because there are response from most participants and generates too few followers to worry about, for many supportive reporting. previous political movements, of miniscule Maintaining high standards will continue to size, have suffered debilitating splits. This be important as long as random selection is toleration for a diversity of views may owe seen as an experimental, “alternative” something to the nature of the subject, namely approach. Voting is not subject to the same that what is being advocated is citizen partici- expectations. When voting fraud occurs, for pation without anyone being able to determine example, the failure is attributed to corrupt who is chosen. But this is speculative. Perhaps individuals or regimes, not to the method of the present state of affairs is due more to the voting itself. In contrast, a major failure of a particular personalities involved, or the policy jury might easily be assumed to be a absence of any rewards of power and money. failure of the whole approach rather than a Would it be incongruous to be rigid about the shortcoming in implementation. implementation of random selection? The test As random selection becomes more widely will come when large amounts of money are used, it is likely to come under attack. Even poured into the process or when policy juries the most well designed operations can be start being used to set government policy. criticized on some ground or other, and it’s Meanwhile, we think it is vital that random also possible to label a process a failure even selection options be subjected to an ongoing when it is entirely successful. In other words, process of critique. New ideas should be doing everything right is no guarantee against encouraged, welcomed, and, when possible, unfair attack. Hence, fear of attack should not tested, yet at the same time all ideas should be be allowed to become too great an inhibition, scrutinized. There simply isn’t enough known otherwise experimentation will come to a halt. about random selection in politics to warrant If random selection is to be developed, some Strategies 83

failures are bound to occur along the way. Should it advocate a range of participatory Indeed, they are important in learning how to alternatives, or focus on one? Should it be do better (so long as there’s plenty of critique). organized locally, nationally, or globally? So here’s another worthwhile tension to There are lots of questions about how a live with: maintaining high standards in using random selection movement might be random selection, yet being willing to experi- organized and what it might do. Surely, if it ment and being prepared to fail sometimes. practices what it preaches, it should use random methods itself. That would be its best Keep the Goal in Mind advertisement, both for members and for outsiders. The task of achieving it lies ahead. Random selection is not the goal itself, but simply a means to a goal, whether it be greater citizen participation, workers’ control, or demarchy. It is important to remain attentive to the ultimate goal — which, of course, can change due to critique, new evidence, or dialogue. For example, the electoral jury, in which citizen panels hear information about candidates for political office and reach conclusions about which ones are preferred, is a valuable means of increasing informed participation in the electoral process. However, if the goal is demarchy — in which there are no elections, only sortition — then the electoral jury is a side track, which may actually serve to legitimize elections and politicians. For promoting demarchy, direct participation in making decisions about the issues, especially by randomly selected citizens as in policy juries, is a more appropri- ate means.

Build a Movement

Many of the beneficial changes in society have been pushed along by social movements, such as those against slavery, for universal literacy, for women’s rights, for workers’ rights, against torture, and for environmental respon- sibility. Representative systems were not introduced by benevolent monarchs, but were the outcome of a complex process that included energetic advocacy. So, let’s set up the “Movement for Random Selection in Politics.” Well, it’s not a very good name, but that can be decided along the way. More importantly, what would be the focus for a movement? Is it for greater use of policy juries? Introduction of demarchy? Appendix Examples of citizen participation

INFORMING Meetings, interviews with elected representatives or bureaucrats Mail out, flyers, noticeboards, newsletters Interviews can be arranged with a govern- ment representative to discuss, say, a Agencies do a direct mail out or a mail box contentious development or political issue. drop or distribute flyers (through key locations Regular interviews can be advertised, too, with such as schools, clubs, shopping malls, or constituents informed of these arrangements in community centers), to their constituents, advance. The process can be largely one-way, advising them of an issue, sometimes seeking with elected representatives or staff members input. The quality can vary according to the wishing to defend their position. Interviews or budget that is available and the skills of the meetings have the capacity to be quite interac- material’s producers. Often this information tive, as a forum for lobbying, a means for won’t make it through the mail box since useful dialogue or worthwhile negotiation. residents increasingly display a sign stating “no junk mail” or “addressed mail only”. Petitions Community noticeboards are used to display information on issues, sometimes seeking Petitions are most often activated by input. All these methods are designed to reach community members or special interest groups the maximum number of people possible. in reaction to a poor decision or in anticipation Regular newsletters are distributed by of an unacceptable decision. They amount to a government agencies or community-based written endorsement of a particular point of organizations via the mail box or other outlets view. Petitions are a means to raise awareness to inform the community of an agency’s and good for giving information to decision activities. Only people with a direct interest in makers as well as establishing a prima facie the agency’s activities would be likely to read case. However, not all signatories are likely to such material. have access to adequate or balanced informa- tion and some could feel pressured to sign. It Press, radio, or television announcements is relatively easy to manipulate the process. Signatures are sought, often at random. Agencies have regular newspaper columns or use the “public notices” section of newspapers Council meetings, parliament (often to fulfill a statutory requirement) to inform the community about issues (e.g., Parliaments and local councils almost always development applications). Agencies also have a public gallery from which constituents publicize their activities on radio or television; can observe those activities that are not held the latter is more likely to be a public relations in-camera (behind closed doors). Some exercise rather than a dissemination of useful councils also offer a public access session in information. Talk-back radio can at least offer which residents can address elected represen- a level of passive consultation as constituents tatives to present their opinions on a proposal. have an opportunity to air their grievances or Excerpts from national parliamentary offer opinions. proceedings (in Australia, for example) are also broadcast on radio or television.

Examples of citizen participation 85

PASSIVE CONSULTATION if a free-call number is used) and feel more personal that a written survey though they Submissions rarely allow for interaction. Hotlines can also be used to create a contact list of those Government bodies are fond of calling for interested in further involvement (Sarkissian submissions, for example when a controversial 1994). proposal is put forward. Submissions can be either written or oral (see public hearings Polls below). The “public notices” section of major daily newspapers indicates the extensive range Opinion polls are used extensively in our of hearings or calls for written submissions. society (see Chapter 2). For example, political This request for public input might come in polls surveying people on their attitudes the early or later stages of decision making. towards various political parties or their Submissions are a means of public involve- policies are common, particularly in the lead- ment though the invitation is most often taken up to an election. Whenever a major issue up by the educated or the articulate. arises, someone will undertake a poll. People have a strong tendency to express appreciation Public hearings for the status quo through opinion polls and then when the status quo is replaced by an Public hearings are the oral equivalent of unwanted alternative, appreciation for the written submissions. In the U.S. they are alternative will emerge (Considine 1994). linked to legislation at all levels of govern- Polls fail to measure the potential for change. ment and so are mandatory and a common Political leaders are increasingly being event. They are designed to gather information condemned for their love of polls and its and opinion necessary to the legislative impact on decisions — this is interpreted as deliberative process. The U.S. experience, “government by market research” or populism. according to critics of public hearings, is that Polls can be conducted either by phone or the process has been hijacked by larger special personal interview and are sometimes incorpo- interest groups and the general public is rarely, rated into telephone hotlines. Usually a yes/no if ever, deeply involved. Hearings risk the or preferential response is all that is required. danger of being controlled by powerful people They are usually randomly selected and a behind large desks using confusing language, statistically significant sample is considered making them largely inaccessible to the typical essential. citizen. Public meetings Telephone hotlines Public meetings are most often held to provide Hotlines are established fairly regularly for a information on an organization’s activities, a whole range of issues. For example, a consult- planned project, or an imminent decision that ant who is conducting the community might be controversial. The emphasis is consultation phase of an environmental study usually on information dissemination rather might set up a hotline for people to register than opinion seeking. People are invited by their views on various options that are being public advertisement. Proceedings are considered. The caller could either find a formalized to allow objectives to be achieved member of staff or a disembodied answering in the limited available time. Public meetings machine at the other end of the phone. This are a good way of providing information to a method of consultation can allow people to large number of people (particularly if visual receive information, register a “vote,” offer a interpretations or displays accompany the suggestion, or lodge a protest. Hotlines are “talking heads”). Public meetings have the easy and cheap for people to use (particularly potential to bring a wide range of people 86 Random selection in politics

together and might incorporate workshops or greatest strength is their accessibility to those panels to create more interaction. Public who might otherwise be excluded (e.g., single officials and community members are cynical parents, people with disabilities, or the aged). about public meetings — probably because They offer a good example of government they tend to attract the incensed and the going to the people. articulate and the process offers little genuine discussion (though small group activities can Listening posts, listening days facilitate this). Often elected representatives use public meetings as a gauge of the Listening posts are a consultative method that importance of an issue — the more people arose from the work of Fran Peavey, an who turn up, the hotter the issue. Community activist who traveled the world, sitting under a members are very resistant to the pervading sign which said “American willing to listen” style of public meetings — being spoken “at” (Peavey 1994). The idea has been used by one or having governing bodies say, “Boy have we of the authors of this book on two occasions got a deal for you”. — one under a sign which read “Candidate willing to listen” prior to a local government Surveys election and on another occasion, post- election, under a banner which read The qualitative equivalent of a quantitative “Councillor willing to listen” (Carson 1996). opinion poll, the survey can involve structured Listening posts offer citizens an opportunity to questioning of a community or a sub-group be heard and work best when the listener which statistically represents the whole avoids a defense of their own actions or population. Surveys can be a way of involving beliefs, instead genuinely listening to citizens’ the public in the early stages of decision concerns that can later be followed up. making. Local people can be recruited to carry Listening days are informal meetings between out surveys if training and income is provided decision makers and those affected by as a way of encouraging community particular decisions and are more focused on ownership of an issue. However there is often specific issues or policies. little discussion and little interaction between participants. Surveys can be very unreliable as Suggestion boxes respondents are prone to give the answers they believe are wanted. One needs to be most Many organizations, such as government careful of bias, design flaws (particularly departments, provide suggestion boxes or superficiality), or unrepresentativeness of customer feedback sheets to give consumers respondents. and citizens an opportunity to make comments about services provided, as well as suggestions Street corner meetings, field trips for how they could be improved. The extent to which suggestions are followed up is variable. Street corner meetings offer a more informal Physical boxes have more recently been and accessible method which allows residents superseded by electronic “boxes” using email. to meet with elected representatives or This can be quick and cheap and avoids paper government staff on the corner of their street use. Electronic suggestions can also lead to whenever a significant issue arises which interaction between participants. One affects them. Residents are advised of the Australian government department found the meeting beforehand. Similarly field trips can process slightly problematic when suggestions be undertaken involving stakeholders and were occasionally offensive and staff members decision makers. Both street corner meetings found analyzing the responses an unpleasant and field trips can combine information task. There need to be structured questions and sharing and information seeking as well as clear guidelines about how suggestions will be mediation or a negotiated outcome. Their processed, evaluated, and acted upon. Examples of citizen participation 87

process. It’s a micro-process that can be Internet and other networks incorporated into a larger consultative process.

A number of government bodies now offer Convergent interviewing information on the internet. Web sites can offer information to constituents as well as Convergent interviewing is another micro- encouraging citizens to interact with the process that combines elements of both agency. Networking between agencies is also structured and unstructured interviews. The occurring. For example, “CouncilNet” is a content is left unstructured but the process is computer network linked to councils highly structured. Convergent interviewing throughout Australia and is designed to help allows for the collection of broad information councils with their environmental management that can be collected quite efficiently. Partici- needs. pants respond to an open-ended question to begin and are then encouraged to talk for a Public rally or street march lengthy period before any specific probe question is asked. Convergent interviewing A public demonstration (rally or march) would can be used (1) to gather information, (2) as seem not to be a “passive” form of consulta- the preliminary stage in deciding which tion, but it belongs in this section because it is questions to ask in a survey, or (3) in deciding not usually interactive with policy makers. It’s which direction to go next with a large-scale more like a physical petition. It’s most often a consultation process. community initiative, arising from frustration with the political process (or lack of consulta- Public access committee tion) and is designed to shame or pressure policy makers into an alternative form of A number of local councils in Australia have action. Rallies can help to create solidarity established public access committees. between like-minded people and spread Consisting of four or five elected representa- awareness about an issue and on occasions tives and staff, these committees meet prior to have proven to be extremely influential in ordinary council meetings and are available changing a government’s direction. for any objectors to discuss contentious issues that are to be determined later by the ACTIVE CONSULTATION governing body.

Delphi Issues forums, workshops, seminars

“Delphi” is an interactive technique designed Usually once-only events, workshops are a to promote participation. Delphi is simply a good way of gauging the level of community series of questionnaires sent to individuals in support for an issue or disseminating informa- order to build consensus. After each round of tion and soliciting community views. They can questionnaires the results are fed back to the involve a diverse group of people; participants participants who are able to change their are usually either self selecting or invited to opinions. Delphi employs the written form attend. They are a good means for testing (though this could also be electronic) in order ideas and can contribute to an overall to focus on ideas rather than personalities. It’s consensus before an action is taken. Skillful inexpensive and flexible. In developing facilitators are needed. On the negative side, consensus it can generate new ideas, dialogue, forums, workshops, and seminars can attract and fruitful discussion among a large or polarized interest groups with entrenched diverse group. It does not suit those who positions and can also be used to manipulate prefer personal contact and can be quite a slow an unsuspecting public with slick presenta- tions and displays. On the positive side, they 88 Random selection in politics

can be an excellent way to encourage proposals or developments. Some precinct community input into agency planning. committees grew out of residents’ action groups, whereas others were established by Advisory committees, working parties more progressive councils. Generally they meet on a regular basis and are often, though Almost all organizations have committees. not always, supported financially. The They’ll vary from short-term working parties financial support is meant to cover admini- to advisory or statutory committees, the latter stration costs and some councils also appoint a appointed by government (see Chapter 4). staff member to co-ordinate the committees’ They are either self-selecting (in response to a activities. They are self-selecting, though there call for interested members), elected, or has been at least one trial of a randomly- nominated. The role of a committee is to selected precinct committee (see Chapter 5). advise an organization on specific issues or Residents’ action groups grow out of a activities and to provide ongoing advice and community’s opposition to government monitoring on community views or specialist decisions. Their role is a lobbying one, issues. They meet regularly and have a formal seeking to apply pressure to government. They structure; they generally are comprised of 15 can lead to considerable community people or less. They keep records of their empowerment and are an excellent training meetings in the form of minutes. They have a ground for activists or elected representatives. tendency to replicate the hierarchies that spawned them. Even when they emerge from Strategic plan an organization that challenges hierarchies, they often can become factionalized with Local councils and government departments tousles for power occurring. Committee complete strategic plans and these increasingly members can weary of the process if the occur with public involvement. Such plans committee does not have a specified “life” or involve the community in an assessment of time frame or purpose. Committees are current and future needs and often provide an frequently reluctant or indifferent to evaluat- opportunity to envision an alternative future. ing their own effectiveness, to determine The result usually involves the setting of whether they are actually achieving their priorities, which are then able to be referred to goals. They offer a good opportunity to gain for major planning decisions. Strategic plans experience in participation and a good channel are notorious for taking up shelf space, i.e. not for community opinion. They can be time being referred to beyond the time of their consuming and have little accountability to the creation and acceptance, but of course this wider community. They are rarely representa- need not be so. Some organizations make tive and can be dominated by members of excellent use of strategic plans in their unequal status. They tend to gather together ongoing decision making. Involvement of the people with quite diverse views and there is community can take many forms and could little potential for resolving issues. They are use a combination of consultative methods, also vulnerable to conscious or unconscious e.g., workshops, surveys, committees, sabotage by critical or negative committee submissions, and so on. Participants are members. usually self-selecting, invited, or nominated.

Precinct committees, residents’ action groups Citizen-initiated referendums

Local councils in Australia have established Citizen-initiated referendums occur when a precinct committees to advise councils on government has determined that a referendum issues which affect their specific geographical can be activated by citizens themselves. A area. Precinct committees can also be used by specified number of signatures is required to councils to provide feedback on shire-wide activate the referendum process; typically the Examples of citizen participation 89

question would go to a referendum at the next addressed by a number of speakers who election. This type of process is used exten- present various opinions. This allows for sively in Switzerland. In the U.S. they are also informed debate by jury members who then known as propositions and have a checkered work towards a consensus. Recommendations though recent history. They are not randomly- are compiled in a report that is referred to selected but all electors ultimately have an policy makers or service providers. The aim is opportunity for involvement. They can lead to to gather together a cross section of the wider short-sighted involvement of people not community, which can be particularly useful directly affected by the decision and the issue for planning or service provision. Ordinary can be reduced to simplistic arguments for and citizens can be reluctant to participate in time against, avoiding informed debate and consuming processes such as these (often discussion (see Chapter 3). With appropriate involving days) and the most remuneration information and education the community’s offered is only barely enough to cover judgment could be soundly based and routine foregone wages. Jury members have evaluated use of referendums would also stimulate the process highly. There is a varying political interest and understanding (as has propensity by elected representatives to use been the case in Switzerland). Referendums the process or to act on juries’ recommenda- need not be either/or — it is possible to offer a tions when the process is actually used. wider choice using a scale to derive a better indication of the range of community views. A Telephone trees major limitation of citizen-initiated referen- dums is when the outcomes are not binding on A telephone tree starts with a telephone call to government. maybe two or three people, advising them of the matter under discussion. Each person is Preferendums then required to ring three or four others to discuss the issue further. This means that a A variation on referendums are preferendums. large number of people can be involved in a A preferendum is a multi-option, decision- very short time in a very interactive way. This making referendum that is designed to begin a consultative method is often used by debate. A preferendum could just as usefully community-based groups and is an effective replace a standard referendum to provide a and efficient form of communication and more considered outcome. Anyone can put information dissemination. It is a quick and forward a suggestion and, during preliminary easy way of tapping into widespread discussions, a list of six to 10 options is drawn community opinion of issues (usually among up which comprehensively reflects the content an existing group of activists). However, as a of the discussions. A preferential system of means of marketing it can be considered voting can be used and a level of consensus on offensive. a complex issue is then expected to become evident. Rather than choosing either/or, voters Search conferences can place their options in order of preference; their preferences will be reflected in the final Search conferences typically run for one or outcome (Emerson 1998). two days and offer an in-depth approach to complex issues (see Chapter 4). The search Citizen panels, planning cells, policy juries process, also known as “future searches”, brings around 30 heterogeneous stakeholders Citizen panels (also called planning cells or together to undertake joint consideration and policy juries) involve, usually, the random planning. Participants are most often invited selection of residents or other stakeholders by the organizers and their selection is related who come together to deliberate on a specific to their level of interest or affectedness. The issue (see Chapter 5). Citizen panels are discussion is sequenced and structured and 90 Random selection in politics

aims to identify a broad cross-section of views assessment. Councils in Australia, however, (Emery & Purser 1996). have begun to develop indicators that will allow an assessment of social impact before Study circles development applications will be considered. The completion of an SIA is the responsibility Study circles involve a small group of people of the developer though it can also be a meeting on a regular basis to discuss ideas that process completed by councils as part of their need more in-depth discussion than shire-wide strategic plans. The public can be community consultation allows for. They can involved via surveys, interviews, or forums, be activated by community members or and existing demographic data are also governments (the latter, for example, in incorporated. Australia in a government initiative in relation to civics education). Study circles have the Electronic voting potential to develop into lobby groups. They can require participants to gather in a single Electronic voting is also known as televoting, location or can occur via computers (the latter and often is associated with electronic town being particular significant for those living in meetings or electronic hearings (see Chapters remote locations). Study circles provide an 4 and 6). Electronic voting usually involves opportunity for information sharing, discus- televised meetings coupled with a phone-in sion, and action. voting facility for a dispersed electorate to express its opinion about an issue. It is Community conventions or many small, short particularly useful for small-scale decision meetings making, for example for a labor union to vote on a motion; this might also involve a video The Kettering Foundation in the U.S. sponsors link-up via satellite. Electronic voting is self a project known as the National Issues Forum. selecting. Participation is limited to those with The meetings are organized through an exten- access to a telephone. sive network of moderators. These 3,000 to 4,000 moderators can be mobilized quickly to Computer conferencing hold meetings on specific issues. The meetings last for only two hours. The participants are Computer conferencing, which allows self-selected but need not be so — randomness instantaneous communication between large could be introduced and the recommendations numbers of people across a country or across could be used to influence political decision the world, involves messages typed into the making. (Crosby forthcoming). participants’ computers to be retrieved by others. The potential of computer conferencing Social impact assessment is for rapid resolution of national problems (albeit superficially) or mass input into large- Local government in Australia is beginning to scale planning for citizens with varying use consultation strategies to assess the social degrees of knowledge and diverse back- impact of major developments. Research is grounds. Again, it is self-selecting but need also occurring throughout the world to find not be so. Participation is limited to those with ways of measuring quality of life or levels of access to a computer. social “capital”, as an alternative or adjunct to existing economic indicators such as gross Youth council domestic product or current account deficit. Social impact assessment is often hindered by A few Australian local councils have an absence of sound indicators that might be established youth councils whose members employed in the same way as environmental result from an informal election among youth indicators are used in environmental impact interest groups. To achieve a cross section, Examples of citizen participation 91

some representatives are required to be elected groups of 500 participants. The small group from indigenous and rural youth, from non- processes ensure that everyone is heard and English speaking backgrounds, school that a diverse group of people represent students, and those who do not attend school. themselves rather than special interest groups. The youth council is then left to young people to organize and is designed to encourage Hypotheticals people to be active in community affairs especially where it involves searching for Hypotheticals are scenarios in which partici- ways and means to improve community life. pants play out roles that simulate a real life The youth council offers recommendations to situation. They are designed to uncover longer the council in its formulation of youth-related term effects of situations that are about to be policies. enacted but whose impact cannot be firmly predicted. For example, the closure of a Design-in, community mapping hospital or a service might lead to unantici- pated outcomes. A hypothetical would When a local or regional area plan is being simulate the situation, participants would be developed, government bodies often use allocated various, and relevant roles, concerns, methods variously called a design-in or and possible outcomes would be uncovered. community mapping. These are interactive approaches that utilize the skills of all PARTNERSHIP participants. They might lead to the develop- ment of environmental, social, or land-use Mediation plans that can be a highly visual, de- emphasizing the written word. Photographs, A number of Australian local councils have illustrated plans, scale representations, craft, introduced mediation policies as a way of scale models, and so on have been used. requiring developers and those who object to Professional planners, architects, designers, their proposals to negotiate their way out of and engineers join together with community conflict. Mediation requires a willingness on members to visualize alternatives, converting both sides to embark on the process. There are their visions into workable plans. The process limitations: it can be seen as a form of social is highly interactive and can exploit the control, as pacifying opponents, creating community’s creative resources, also allowing unrealistic expectations if the parties are poles less articulate people a “voice.” apart, forfeiting important principles and values if the resolution is not binding. There Charrette are also strengths in the process: mediation can break down stereotypes (e.g., of develop- Similar to a design-in, a charrette (from the ers or greenies) and be replaced by a French word meaning “little cart”) has come recognition of people as real people; the to mean “feverish work to meet a deadline community can be empowered by a process of with some public input”. A charrette brings negotiation which is usually denied them; together a disparate group of community mediation can provide creative options to help members and government officials in order to solve vexatious problems; and it can be a reach consensus. Using small group and godsend for staff and elected representatives plenary sessions, the participants work on when criticisms are deflected away from them. aspects of the problem, working and reporting The process is inevitably self selecting, but if back within a prescribed deadline. Designers, the issue involves a large enough group, there planners, and architects collaborate with is no reason why random selection could not community members on the proposal until be employed. everyone reaches agreement and a workable proposal is assured. They can be very large 92 Random selection in politics

Referendums Consensus conferences

Like citizen-initiated referendums (see above Consensus conferences (see Chapter 4) are a and also Chapter 3), this statutory method has participatory, deliberative approach to policy the potential to involve all voters. Referen- making or problem solving usually involving dums are usually binding if a proportion of the technology assessment. In some instances (for population supports them (which may be more example in Denmark), their recommendations than 50%). Referendums represent a genuine are discussed in parliaments, thereby decision making partnership between qualifying this method as a potential example government and the general public and are a of partnership. Participants are most often self- good example of direct democracy. However, selecting or chosen by organizers; random they lack a strong, deliberative component. selection can be used. Organizers follow procedures associated with citizen panels: a Community management committees small group (10-16) hears evidence, asks questions, discusses and reflects, then makes There are some rare instances of community recommendations. Consensus conferences management committees being given decision usually run over at least three days. making powers and resources to support their Participants can vary the agenda and write decisions. The committee could be comprised their own report. They have now been used in of either elected or randomly selected numerous countries in a limited way. community members and would have delegated authority to spend within a specified Combination of methods budget. In a couple of Australian local councils these committees are allocated a A combination of the above methods is proportion of Council’s income from rates to possible and there have been some impressive be spent on their local area. It can be very examples of this: a three-step procedure empowering for a local community to manage developed by Renn et al. (see Chapter 6) its own affairs in this way. A good partnership combines Delphi, citizens panels, visioning, between the community and bureaucrats is and so on. A citizen survey panel directed by required and a genuine commitment by elected Lyn Kathlene (see Chapter 6) combined representatives to share power with the surveys, interviews, an advisory committee, constituents they represent. etc. The combinations are limited only by political will, determination, and the creativity Social contracts of all actors in the participatory process.

There have been attempts to negotiate social contracts, for example between large develop- ers and local authorities (Club Med and a coastal council in Australia, Byron Shire, provides one case study of this method). A controversy can be resolved by the drawing up of a social contract between the proponents, the opponents, and the governing body. These contracts are not legally binding but represent a commitment by a developer to satisfy certain community needs which may not be within the statutory power of the local council to apply or enforce. The details are negotiated between all parties.

Bibliography

Abramson, Jeffrey. We, the Jury: The Jury Barnes, Barry. Scientific Knowledge and System and the Ideal of Democracy. New Sociological Theory. London: Routledge York: BasicBooks, 1994. and Kegan Paul, 1974. Abramson, Jeffrey B., F. Christopher Arterton, Becker, Ted, Paul Szep, with Dwight Ritter. and Gary R. Orren. The Electronic Un-Vote for a New America (A Guide to Commonwealth: The Impact of New Media Constitutional Revolution). Boston, MA: Technologies on Democratic Politics. New Allyn and Bacon, 1976. York: Basic Books, 1988. Becker, Theodore L. American Government: Alcoff, Linda. “The problem of speaking for Past•Present•Future. Boston, MA: Allyn others.” Cultural Critique, no. 20 (1991): and Bacon, 1976. 5-32. ______. Teledemocracy Action News + Anweiler, Oskar. The Soviets: The Russian Network. The WebSite of the Global Workers, Peasants, and Soldiers Councils, Democracy Movement, 1905-1921. New York: Pantheon, 1974. , 1998. mondsworth, UK: Penguin, 1984. Bookchin, Murray. “What is Communalism? Arnstein, Sherry R. “A Ladder of Citizen The Democratic Dimension of Anarchism.” Participation.” AIP Journal, July (1969): Green Perspectives, no. 31, October 1994, 216-224. pp. 1-6. Arterton, F. Christopher. Teledemocracy: Can Boserup, Anders, and Andrew Mack. War Technology Protect Democracy? Newbury Without Weapons: Non-Violence in Park, CA: Sage, 1987. National Defense. London: Frances Pinter, Asher, Herbert. Polling and the Public: What 1974. Every Citizen Should Know. 3rd ed. Broome, John. “Selecting People Randomly.” Washington, DC: CQ Press, 1995. Ethics, vol. 95 (October 1984): 38-55. Atlee, Tom. “Transformational Politics.” In Brown, Colin. “Greater Democracy, Better Context, no. 30 (Fall 1991): 56-58. Decisions.” Consumer Policy Review, vol. Aubert, Vilhelm. “Chance in Social Affairs.” 7, no. 5 (1997): 170-173. Inquiry, vol. 2 (1959): 1-24. Burnheim, John. Is Democracy Possible? The Avery, Michael, Brian Auvine, Barbara Alternative to Electoral Politics. London: Streibel, and Lonnie Weiss. Building Polity Press, 1985. United Judgment: A Handbook for Consen- ______. “Democracy, Nation States and the sus Decision Making. Madison, WI: Center World System.” In New Forms of Democ- for Conflict Resolution, 1981. racy, edited by David Held and Christopher Bakalar, James B., and Lester Grinspoon. Pollitt, 218-239. London: Sage, 1986. Drug Control in a Free Society. Cam- ______. “Democracy by Statistical Represen- bridge, UK: Cambridge University Press, tation.” Social Alternatives, vol. 8, no. 4 1984. (January 1990): 25-28. Barber, Benjamin R. Strong Democracy: ______. “Power-Trading and the Environ- Participatory Politics for a New Age. ment.” Environmental Politics, vol. 4, no. 4 Berkeley, CA: University of California (Winter 1995): 49-65. Press, 1984. Burrowes, Robert J. The Strategy of Nonvio- ______. “Opinion Polls: Public Judgment or lent Defense: A Gandhian Approach. Private Prejudice?” The Responsive Albany, NY: State University of New York Community, vol. 2, no. 2 (1992): 4-6. Press, 1996. 94 Random selection in politics

Bussemaker, Jet, and Rian Voet. “Citizenship Considine, Mark. Public Policy: A Critical and Gender: Theoretical Approaches and Approach. Melbourne, Australia: Historical Legacies.” Critical Social Macmillan Education Australia, 1994. Policy, vol. 18, no. 3 (1998): 277-307. Coote, Anna, and Jo Lenaghan. Citizens’ Butler, David, and Austin Ranney, eds. Juries: Theory into Practice. London: Referendums Around the World: The Institute for Public Policy Research, 1997. Growing Use of Direct Democracy. Coover, Virginia, Ellen Deacon, Charles Washington, DC: American Enterprise Esser, and Christopher Moore Resource Institute Press, 1994. Manual for a Living Revolution. Philadel- Callenbach, Ernest, and Michael Phillips. A phia, PA: New Society Publishers, 1981. Citizen Legislature. Berkeley, CA: Banyan Crain, Robert L., Elihu Katz, and Donal B. Tree Books, 1985. Rosenthal. The Politics of Community Campbell, Blair. “Paradigms Lost: Classical Conflict: The Fluoridation Decision. Athenian Politics in Modern Myth.” Indianapolis, IN: Bobbs-Merrill, 1969. History of Political Thought, vol. 10, no. 2 Cronin, Thomas E. Direct Democracy: The (Summer 1989): 189-213. Politics of Initiative, Referendum, and Carson, Lyn. “The Jury is IN: Parent Juries as Recall. Cambridge, MA: Harvard Univer- an Empowerment Tool in Education.” sity Press, 1989. Community Quarterly, no. 33 (1994): 18- Crosby, Ned. “The Peace Movement and New 20. Democratic Processes.” Social Alternatives, ______. “How do decision makers in local vol. 8, no. 4 (1990): 33-37. government respond to public participa- ______. Citizens Election Forum. Forthcom- tion? Case study: Lismore City Council ing. 1991-1995.” Unpublished PhD thesis. Crosby, Ned, Janet M. Kelly, and Paul Lismore, Australia: Southern Cross Univer- Schaefer. “Citizens Panels: A New sity, 1996. Approach to Citizen Participation.” Public ______. “Converting good ideas into sound Administration Review, vol. 46, no. 2 policy and practice: community consulta- (1986): 170-178. tion in local government.” Paper presented Dahl, Robert A. After the Revolution? at the Reaching Common Ground - Open Authority in a Good Society. New Haven, Government Network Conference, CT: Yale University Press, 1970. Brighton-le-Sands, 1998. ______. A Preface to Economic Democracy. ______, and Ross Roache. Consultation and Berkeley, CA: University of California Participation: Study Module 2 "Methods". Press, 1985 Lismore, Australia: Southern Cross Univer- ______, and Edward R. Tufte. Size and sity, 1996. Democracy. Stanford, CA: Stanford Coates, Ken.Work-ins, Sit-ins and Industrial University Press, 1973. Democracy: The Implications of Factory Dator, Jim. “The 1982 Honolulu Electronic Occupations in Great Britain in the Early Town Meeting.” In The Future of Politics: ’Seventies. Nottingham, UK: Spokesman, Governance, Movements and World Order, 1981. edited by William Page, 211-220. London: Cole, Leonard A. “Resolving Science Contro- Frances Pinter, 1983. versies: From Science Court to Science Davies, Alan. “Education in a Public Service Hearings Panel.” In Governing Science and Union. A Strategy for Adaption.” In Technology in a Democracy, edited by Canberra Papers in Continuing Education Malcolm L. Goggin. Knoxville, TN: The New Series 2, edited by Nicolas Haines, 23- University of Tennessee Press, 1986. 39. Canberra, Australia: Centre for Connolly, William E. The Terms of Political Continuing Education, Australian National Discourse. 2nd ed. Oxford, UK: Martin University, 1982. Robertson, 1983. Bibliography 95

______. “Organisational Change in a Trade Decision-Making and Electoral Systems. Union.” Work and People, vol. 11, no. 1 Belfast, Northern Ireland: The De Borda (1985): 10-19. Institute, 1998. Davis, Morton D. Game Theory: A Nontechni- Emery, Fred E. Toward Real Democracy and cal Introduction. New York: Basic Books, Toward Real Democracy: Further 1970. Problems, Occasional Paper. Toronto, Denver, David, Gordon Hands, and Bill Jones. Canada: Ontario Ministry of Labour, “Fishkin and the Deliberative Opinion Poll: 1989a. Lessons from a Study of the Granada 500 ______. “Industry Councils — Comments on Television Program.” Political Communi- One Aspect of the Jackson Report.” In cation, vol. 12, no. 2 (April 1995): 147-156. Participative Design for Participative Dienel, Peter C. Die Planungszelle: Eine Democracy, edited by Merrelyn Emery, Alternative zur Establishment-Demokratie. 160-166. Canberra, Australia: Centre for 2nd ed. Opladen, Germany: Westdeutscher Continuing Education, Australian National Verlag, 1988. University, 1989b. Dienel, Peter C. “Contributing to Social ______. “The Jury System and Participative Decision Methodology: Citizen Reports on Democracy.” In Participative Design for Technological Projects.” In Social Decision Participative Democracy, edited by Methodology for Technological Projects, Merrelyn Emery, 167-172. Canberra, edited by Charles Vlek and George Australia: Centre for Continuing Education, Cvetkovich, 133-151. Dordrecht, Nether- Australian National University, 1989c. lands: Kluwer, 1989. ______, and Einar Thorsrud. Democracy at Dienel, Peter C., and Ortwin Renn. “Planning Work. The Report of the Norwegian Cells: A Gate to ‘Fractal’ Mediation.” In Industrial Democracy Program, Leiden, Fairness and Competence in Citizen Netherlands: Martinus Nijhoff Social Participation: Evaluating Models for Sciences Division, 1976. Environmental Discourse, edited by Ortwin Emery, Merrelyn, and Ronald E. Purser. The Renn, Thomas Webler and Peter Search Conference: A Powerful Method for Wiedemann, 117-140. Dordrecht, Planning Organizational Change and Netherlands: Kluwer Academic, 1995. Community Action. San Francisco, CA: Dolgoff, Sam, ed. The Anarchist Collectives: Jossey-Bass, 1996. Workers’ Self-Management in the Spanish Engelstad, Fredrik. “The Assignment of Revolution, 1936-1939. New York: Free Political Office by Lot.” Social Science Life Editions, 1974. Information, vol. 28, no. 1 (March 1989): Domitrovich, Stephanie. “Jury Source Lists 23-50. and the Community’s Need to Achieve Enright, Seán, and James Morton. Taking Racial Balance on the Jury.” Duquesne Liberties: The Criminal Jury in the 1990s. Law Review, vol. 33, no. 1 (1994): 39-103. London: Weidenfeld and Nicolson, 1990. Edelman, Murray. Politics as Symbolic Erez, Edna. “Random Assignment, the Least Action: Mass Arousal and Quiescence. Fair of Them All: Prisoners’ Attitudes Chicago, IL: Markham, 1971. Toward Various Criteria of Selection.” Eliasoph, Nina. Avoiding Politics: How Criminology, vol. 23, no. 2 (May 1985): Americans Produce Apathy in Everyday 365-379. Life. Cambridge, UK: Cambridge Univer- Feyerabend, Paul K. Against Method: Outline sity Press, 1998. of an Anarchistic Theory of Knowledge. Ellickson, Robert C. Order Without Law: How London: New Left Books, 1975. Neighbors Settle Disputes. Cambridge, Fienberg, Stephen E. “Randomization and MA: Harvard University Press, 1991. Social Affairs: The 1970 Draft Lottery.” Emerson, Peter J. Beyond the Tyranny of the Science, vol. 171 (22 January 1971): 255- Majority. Voting Methodologies in 261. 96 Random selection in politics

Finkel, Norman J. Commonsense Justice: Ginsberg, Benjamin. The Consequences of Jurors’ Notions of the Law. Cambridge, Consent: Elections, Citizen Control and MA: Harvard University Press, 1995. Popular Acquiescence. Reading, MA: Fishkin, James S. Democracy and Delibera- Addison-Wesley, 1982. tion: New Directions for Democratic ______. The Captive Public: How Mass Reform. New Haven, CT: Yale University Opinion Promotes State Power. New York: Press, 1991. Basic Books, 1986. ______. “The Televised Deliberative Poll: Gittell, Marilyn. Limits to Citizen Participa- An Experiment in Democracy.” Annals of tion: The Decline of Community Organiza- the American Academy of Political and tions. Beverly Hills, CA: Sage, 1980. Social Science, vol. 546 (July 1996): 132- Goodwin, Barbara. Justice by Lottery. 140. Chicago, IL: University of Chicago Press, ______. The Voice of the People: Public 1992. Opinion and Democracy. New Haven, CT: Gould, Carol C. Rethinking democracy. Yale University Press, expanded edition, Freedom and social cooperation in politics, 1997. economics, and society. New York: Fixdal, Jon. “Consensus Conferences as Cambridge University Press, 1988. ‘Extended Peer Groups’.” Science and Gramsci, Antonio. Prison Notebooks. New Public Policy, vol. 24, no. 6 (1997): 366- York: International Publishers, 1971. 376. Grossman, Lawrence K. The Electronic Frey, Bruno S., and Werner W. Pommerehne. Republic: Reshaping Democracy in the “On the fairness of pricing — an empirical Information Age. New York: Viking survey among the general population.” Penguin, 1995. Journal of Economic Behavior and Organi- Guérin, Daniel. Anarchism: From Theory to zation, vol. 20, no. 3 (April 1993): 295- Practice. New York: Monthly Review 307. Press, 1970. Fukurai, Hiroshi, Edgar W. Butler, and Habermas, Jurgen. Toward a Rational Society. Richard Krooth. “Cross-Sectional Jury Translated by Jeremy J. Shapiro, London: Representation or Systematic Jury Repre- Heinemann, 1971. sentation? Simple Random and Cluster Hans, Valerie P., and Neil Vidmar. Judging Sampling Strategies in Jury Selection.” the Jury. New York: Plenum, 1986. Journal of Criminal Justice, vol. 19, no. 1 Hansen, Mogens Herman. The Athenian (1991): 31-48. Democracy in the Age of Demosthenes: Garbe, Detlef. “Planning Cell and Citizen Structure, Principles and Ideology. Oxford, Report: A Report on German Experiences UK: Basil Blackwell, 1991. with New Participation Instruments.” Headlam, James Wycliffe. Election by Lot at European Journal of Political Research, Athens. Cambridge, UK: Cambridge vol. 14 (1986): 221-236. University Press, [1891] 1933. Gastil, John. Democracy in Small Groups: Hesse, Mary. The Structure of Scientific Participation, Decision Making, and Inference. London: Macmillan, 1974. Communication. Philadelphia, PA: New Hirst, Paul. “Is Democracy Possible?” (book Society Publishers, 1993. review). Sociological Review, vol. 34, no. 3 Gaventa, John. Power and Powerlessness: (August 1986): 669-673. Quiescence and Rebellion in an Appala- Hunnius, Gerry, David G. Garson, and John chian Valley. Oxford, UK: Oxford Case. eds. Workers’ Control: A Reader on University Press, 1980. Labor and Social Change. New York: Gilliom, John. Surveillance, Privacy, and the Vintage, 1973. Law: Employee Drug Testing and the Jacobsohn, Gary J. “Citizen Participation in Politics of Social Control. Ann Arbor, MI: Policy-Making: The Role of the Jury.” University of Michigan Press, 1994. Bibliography 97

Journal of Politics, vol. 39, no. 1 (February Krimsky, Sheldon. “Local Control of Research 1977): 73-96. Involving Chemical Warfare Agents.” In Jefferson Center, Policy Jury on School-Based Governing Science and Technology in a Clinics: Final Report. Minneapolis, MN: Democracy, edited by Malcolm L. Goggin. Jefferson Center, August 1988. Knoxville, TN: The University of Jones, A. H. M. Athenian Democracy. Oxford, Tennessee Press, 1986. UK: Blackwell, 1960. Laffin, Martin, and Martin Painter. eds. Joss, Simon. “Danish Consensus Conferences Reform and Reversal. Lessons from the as a Model of Participatory Technology Coalition Government in NSW 1988-1995. Assessment: An Impact Study of Consen- Melbourne, Australia: Macmillan, 1995. sus Conferences on Danish Parliament and Landry, Charles, David Morley, Russell Danish Public Debate.” Science and Public Southwood, and Patrick Wright. What a Policy, vol. 25, no. 1 (1998): 2-22. Way to Run a Railroad: An Analysis of Joss, Simon, and John Durant, eds. Public Radical Failure. London: Comedia, 1985. Participation in Science. London: Science Lane, Mary. “Community Development and a Museum & European Commission Direc- Postmodernism of Resistance.” In Emanci- torate General X11, 1995. patory Practice in Social Work: Critical Jump, James W. “The Only Fair Way for Elite Postmodern Perspectives, edited by Bob Colleges to Choose Their Freshman Pease and Jan Fook. Sydney, Australia: Classes Is by Random Selection.” Chroni- Allen & Unwin, forthcoming. cle of Higher Education (27 April 1988): Langbein, John H. “On the Myth of Written A52. Constitutions: The Disappearance of Kalven, Harry Jr., and Hans Zeisel. The Criminal Jury Trial.” Harvard Journal of American Jury. Chicago, IL: University of Law & Public Policy, vol. 15, no. 1 (Winter Chicago Press, 1971. 1992): 119-127. Kaner, Sam. Facilitator’s Guide to Participa- Luce, Robert Duncan, and Howard Raiffa. tory Decision-Making. Philadelphia, PA: Games and Decisions: Introduction and New Society Publishers, 1996. Critical Survey. New York: Wiley, 1957. Kantowsky, Detlef. Sarvodaya: The Other Lukes, Steven. Power: A Radical View. Development. New Delhi, India: Vikas, London: Macmillan, 1974. 1980. Lummis, C. Douglas. Radical Democracy. Kathlene, Lyn, and John A. Martin. “Enhanc- Ithaca, NY: Cornell University Press, 1996. ing Citizen Participation: Panel Designs, Lynch, Tony. “Debating Democracy.” Perspectives, and Policy Formation.” Economy and Society, vol. 18, no. 1 Journal of Policy Analysis and Manage- (February 1989): 110-124. ment, vol. 10, no. 1 (1991): 46-63. McGuire, Rick. “Athletes at Risk.” In Athletes Kipnis, David. The Powerholders. 2nd ed. at Risk: Drugs and Sport, edited by Ray Chicago, IL: University of Chicago Press, Tricker, and David L. Cook, 1-14. 1981. Dubuque, IA: Brown, 1990. ______. Technology and Power. New York: Mandela, Nelson. Long Walk To Freedom: Springer-Verlag, 1990. The Autobiography of Nelson Mandela. Knag, Sigmund. “Let’s Toss for It: A London: Abacus, 1995. Surprising Curb on Political Greed.” Manin, Bernard. The Principles of Represen- Independent Review, vol. 3, no. 2 (Fall tative Government. Cambridge, UK: 1998): 199-209. Cambridge University Press, 1997. Kohr, Leopold. The Breakdown of Nations. Mansbridge, Jane J. “Time, Emotion, and London: Routledge and Kegan Paul, 1957. Inequality: Three Problems of Participatory Kotler, Milton. Neighborhood Government: Groups.” Journal of Applied Behavioral The Local Foundations of Political Life. Science, vol. 9, no. 2/3 (1973): 351-370. Indianapolis, IN: Bobbs-Merrill, 1969. 98 Random selection in politics

______. Beyond Adversary Democracy. New Oppenheim, Felix E. “Equality, Groups, and York: Basic Books, 1980. Quotas.” American Journal of Political Martin, Brian. Social Defense, Social Change. Science, vol. 21, no. 1 (February 1977): 65- London: Freedom Press, 1993. 69. ______. “Democracy Without Elections.” Orren, Gary. “Fall from Grace: The Public’s Social Anarchism, no. 21 (1995): 18-51. Loss of Faith in Government.” In Why Martin, Thomas S. “Unhinging All Govern- People Don’t Trust Government, edited by ment: The Defects of Political Representa- Joseph S. Nye, Jr., Philip D. Zelikow, and tion.” Our Generation, vol. 20, no. 1 (Fall David C. King, 77-107. Cambridge, MA: 1988): 1-21. Harvard University Press, 1997. Merkle, Daniel M. “The National Issues Pateman, Carole. Participation and Democ- Convention Deliberative Poll.” Public ratic Theory. Cambridge, UK: Cambridge Opinion Quarterly, vol. 60, no. 4 (Winter University Press, 1970. 1996): 588-619. Payne, Stanley L. The Art of Asking Questions. Michels, Robert. Political Parties: A Princeton, NJ: Princeton University Press, Sociological Study of the Oligarchical 1951. Tendencies of Modern Democracies. New Peavey, Fran. Heart Politics. Montréal, York: Dover, [1915] 1959. Canada: Black Rose Books, 1986. Morgan, Edmund S. Inventing the People: The ______. By Life's Grace, Musings on the Rise of Popular Sovereignty in England Essence of Social Change. Philadelphia, and America. New York: Norton, 1988. PA: New Society, 1994. Morris, David, and Karl Hess. Neighborhood Phillips, Anne. Engendering Democracy. Power: The New Localism. Boston, MA: Cambridge and Oxford, UK: Polity Press, Beacon, 1975. 1991. Mottram, D. R. “Introduction — Drugs and ______. The Politics of Presence. Oxford, Their Use in Sport.” In Drugs in Sport, UK: Clarendon Press, 1995. edited by D.R. Mottram, 1-31. London: Piven, Frances Fox, and Richard A. Cloward. Spon, 1988. Why Americans Don’t Vote. New York: Moyer, Bill. The Practical Strategist: Pantheon, 1988. Movement Action Plan (MAP). Strategic Price, Vincent, and Peter Neijens. “Delibera- Theories for Evaluating, Planning, and tive Polls: Toward Improved Measures of Conducting Social Movements. San ‘Informed’ Public Opinion?” International Francisco, CA: Social Movement Journal of Public Opinion Research, vol. Empowerment Project, 1993. 10, no. 2 (Summer 1998): 145-176. Mueller, Charles E. “A Model ‘Antitrust Bill Randle, Michael. Civil Resistance. London: of Rights?’ The Citizen Jury in World Fontana, 1994. Antimonopoly Policy.” Antitrust Law & Rapoport, Anatol. Fights, Games, and Economics Review, vol. 28, no. 4 (1997): Debates. Ann Arbor, MI: University of 73-90. Michigan Press, 1960. Mueller, Dennis C., Robert D. Tollison, and Raptis, Michael. Revolution and Counter- Thomas D. Willett. “Representative Revolution in Chile: A Dossier on Workers’ Democracy via Random Selection.” Public Participation in the Revolutionary Process. Choice, vol. 12 (Spring 1972): 57-68. London: Allison & Busby, 1974. Mulgan, Richard G. “Lot as a Democratic Ratna, Anurug. “Sarvodaya democracy.” Device of Selection.” Review of Politics, Social Alternatives, vol. 8, no. 4 (1990): vol. 46, no. 4 (1984): 539-560. 38-42. Munsterman, G. Thomas, and Janice T. Renn, O. “Decision Analytic Tools for Munsterman. “The Search for Jury Repre- Resolving Uncertainty in the Energy sentativeness.” Justice System Journal, vol. Debate.” Nuclear Engineering and Design, 11, no. 1 (Spring 1986): 59-78. vol. 93 (1986): 167-179. Bibliography 99

______, H. U. Stegelmann, G. Albrecht, U. Sclove, Richard E. Democracy and Technol- Kotte, and H. P. Peters. “An Empirical ogy. New York: Guilford Press, 1995. Investigation of Citizens’ Preferences Seiler, Hans-Jorg. “Review of "Planning Among Four Energy Scenarios.” Techno- Cells:" Problems of Legitimation.” In logical Forecasting and Social Change, Fairness and Competence in Citizen vol. 26 (1984): 11-46. Participation: Evaluating Models for ______, Thomas Webler, Horst Rakel, Peter Environmental Discourse, edited by Ortwin Dienel, and Branden Johnson. “Public Renn, Thomas Webler and Peter Participation in Decision Making: A Three- Wiedemann, 141-156. Dordrecht, Nether- Step Procedure.” Policy Sciences 26 lands: Kluwer Academic, 1995. (1993): 189-214. Simon, Rita James. ed. The Jury System in ______, ______, and Peter Wiedemann, eds. America: A Critical Overview. Beverly Fairness and Competence in Citizen Hills, CA: Sage, 1975. Participation: Evaluating Models for Sinclair, R. K. Democracy and Participation Environmental Discourse. Dordrecht, in Athens. Cambridge, UK: Cambridge Netherlands: Kluwer, 1995. University Press, 1988. Richards, Vernon. Lessons of the Spanish Sharp, Gene. The Politics of Nonviolent Revolution (1936-1939). 3rd ed. London: Action. Boston, MA: Porter Sargent, 1973. Freedom Press, 1983. Sharp, Gene, with Bruce Jenkins. Civilian- Ritzer, George. The McDonaldization of based Defense: A Post-Military Weapons Society. rev. ed. Thousand Oaks, CA: Pine System. Princeton, NJ: Princeton University Forge Press, 1996. Press, 1990. Roberts, Ernie. Workers’ Control. London: Slaton, Christa Daryl. Televote: Expanding Allen & Unwin, 1973. Citizen Participation in the Quantum Age. Root and Branch eds. Root & Branch: The New York: Praeger Publishers, 1992. Rise of the Workers’ Movements. Soros, George (Geoff Shandler, ed). The Greenwich, CT: Fawcett, 1975. Crisis of Global Capitalism: Open Society Ryan, Howard. Blocking Progress: Consensus Endangered. Public Affairs, 1998 Decision Making in the Anti-Nuclear Sproull, Lee, and Sara Kiesler. Connections: Movement. Berkeley, CA: Overthrow New Ways of Working in the Networked Cluster, Livermore Action Group, 1985, Organization. Cambridge, MA: MIT Press, published in revised form as chapter 18 in 1991. Ryan, Howard. Critique of Nonviolent Steele, David Ramsay. “Why Stop at Term Politics: From Mahatma Gandhi to the Limits? Would the First 535 Names in the Anti-Nuclear Movement. Phone Book, or any other Randomly , 1996. Selected Group of People, Be an Improve- Sale, Kirkpatrick. Human Scale. New York: ment over Our Current Representatives?” Coward, McCann and Geoghegan, 1980. National Review, vol. 47, no. 17 (11 Salomon, Jean-Jacques. Science and Politics. September 1995): 38-42. London: Macmillan, 1973. Stern, Philip M. The Best Congress Money Sarkissian, Wendy. “Community Participation Can Buy. New York: Pantheon, 1988. in Theory and Practice.” In Community Sudman, Seymour, and Norman M. Bradburn. Participation in Practice: Casebook, edited Asking Questions. San Francisco, CA: by Wendy Sarkissian and Kelvin Walsh, 1- Jossey-Bass, 1982. 32. Perth: Institute for Science and ______, ______, and Norbert Schwarz. Technology Policy, Murdoch University, Thinking About Answers: The Application 1994. of Cognitive Processes to Survey Method- Schmidt, David D. Citizen Lawmakers: The ology. San Francisco, CA: Jossey-Bass, Ballot Initiative Revolution. Philadelphia, 1996. PA: Temple University Press, 1989. 100 Random selection in politics

Theobald, Robert. “The Healing Century.” ABC Radio National Transcripts, , 12 April 1998. Thornley, Jenny. Workers’ Cooperatives: Jobs and Dreams. London: Heinemann, 1981. Toft, Jesper. “Denmark: Seeking Broad-Based Consensus on Gene Technology.” Science and Public Policy, vol. 23, no. 3 (1996): 171-174. Tsagarousianou, Roza, Damian Tambini, and Cathy Bryan. Cyberdemocracy: Technol- ogy, Cities and Civic Networks. London: Routledge, 1998. Wadler, Gary I., and Brian Hainline. Drugs and the Athlete. Philadelphia, PA: Davis, 1989. Watner, Carl, “‘Stateless, Not Lawless’: Voluntaryism and Arbitration,” The Voluntaryist, no. 84 (February 1997): 1-8. Wernick, Andrew. Promotional Culture: Advertising, Ideology and Symbolic Expression. London: Sage, 1991. Wolfle, Dael. “Chance, or Human Judgment?” Science, vol. 167 (27 February 1970): 168. Woodford, James. “First Prize in the Loggers’ Lottery: You’re Sacked.” Sydney Morning Herald (2 December 1998): 1. Wortman, Camille B., and Rabinowitz, Vita C. “Random Assignment: The Fairest of Them All.” Evaluation Studies Review Annual, vol. 4 (1979): 177-184. Zerman, Melvyn Bernard. Beyond a Reason- able Doubt: Inside the American Jury System. New York: Thomas Y. Crowell, 1981.