Deliberative Democracy in an Unlikely Place: Deliberative Polling in China

James S. Fishkin, Stanford University Baogang He, Deakin University Robert C. Luskin, University of Texas at Austin Alice Siu, Stanford University

Talk of democratic reform sometimes focuses on talk itself. The idea animating deliberative democracy is that the mass public can make a substantive contribution to public policy by engaging in a certain kind of public discussioncitizen deliberation. The common presumption is that this is an advanced variant of democracy and thus possible only in the most thoroughly democratic countries. The contemporary literature commonly treats it as a democratic component of a Rawlsian just society or as a Habermasian ideal that can be approached only under the most favorable conditions. 1 Other treatments are equally aspirational, attempting to build on perfected versions of civic Republicanism or Liberalism, for societies ordered by public reason. 2 There is some question, indeed, whether deliberative democracy is possible at all, even in the United States and other long-standing democracies. Richard Posner, for example, has dismissed it as a pipe dream hardly worth the attention of a serious person. Focusing on citizen deliberation as a way of improving democracy, he says, would be like asking Odysseus to sprout wings as a way of leaving Calypsos island. What makes deliberation unattainable, in his view, is

For a lucid overview making clear this aspirational character, see Joshua Cohen Deliberation

and Democratic Legitimacy in James Bohman and William Rehg, eds., Deliberative Democracy: Essays on Reason and Politics (Cambridge, MA: MIT Press, 1997). See his section I. See also Cass R. Sunstein Health Trade-offs in Jon Elster, ed., Deliberative Democracy (Cambridge: Cambridge University Press, 1998 ), pp 232-259, who holds that the aspiration to achieve deliberative democracy is a defining feature of U.S. constitutionalism p. 232.
2

For a good summary followed by major essays, see introduction, Bohman and Rehg.

citizens inability to deal with complex policy issues. 3 Thus democratic reforms should focus simply on improving the conditions for party competition. Other skeptics claim that the deliberative ideal is unattainable because real-world deliberations will be distorted by inequalities of gender or class 4 , or will replicate a predictable pattern of polarization rather than a real consideration of issues on their merits. 5 Are the skeptics correct? Concrete applications of deliberative democracy raise three basic questions: a) Who: who is doing the talking? How are they selected? To what extent are they a representative sample of some larger population? b) How: To what extent and in what ways does the discussion satisfy deliberative and democratic aspirations?

Richard A. Posner, Law, Pragmatism and Democracy (Cambridge: Harvard University Press,

2004), p. 163.
4

See Lynn M. Sanders, Against Deliberation Political Theory Vol 25, no 3 (June 1997), 347-

376 and Iris Marion Young Intersecting Voices: Dilemmas of Gender, Political Philosophy and Policy (Princeton: Princeton University Press, 1997), chapter III and Inclusion and Democracy (Oxford: Oxford University Press, 2000 ), chapter two.
5

See Cass R. Sunstein Deliberative Trouble? Why Groups Go to Extremes, Yale Law Journal,

110, (2000), 71-119 and The Law of Group Polarization in James S. Fishkin and Peter Laslett, eds., Debating Deliberative Democracy (Oxford: Blackwell, 2003) and also Republic.Com (Princeton University Press, 2002), chapter 3.

c) To what effect: What difference does the discussion make? To what extent do the results have sufficient legitimacy and connections to the policy process to be implemented? Can these questions be answered affirmatively? If so, where? This paper describes a local deliberative experiment in China. If deliberative democracy is unattainable even in advanced democracies, one might think it even further out of reach in modestly democratized or partially authoritarian systems like Chinas. Yet the process and results we describe offer plausibly affirmative responses to all three questions. They constitute a kind of possibility proof, showing what might be attainable, even in a country like China, under favorable conditions. Deliberative Democracy and Public Consultation Throughout the world, policy makers who wish to consult the public face an apparent dilemma. On the one hand, if they consult mass opinion directly, they will get views that are largely uninformed. Most citizens, most of the time, in most political systems, know little about the details of public policy. 6 But policy elites and organized interests, which may have different
6

See Philip Converse The Nature of Belief Systems in Mass Publics in David E. Apter, ed.,

Ideology and Discontent (New York: Free Press, 1964), Robert C. Luskin, Measuring Political Sophistication, American Journal of Political Science 31 (November, 1987): 856-99; Michael X. Delli Carpini and Scott Keeter, What Americans Know about Politics and Why It Matters (New Haven: Yale University Press, 1996); Donald M. Kinder, Opinion and Action in the Realm of Politics in Daniel T. Gilbert, Susan T. Fiske, and Gardner Lindzey, eds., The Handbook of Social Psychology, 4th ed. (Boston: McGraw-Hill, 1998); and Vincent Price,

interests and values, may not speak accurately for the people. We seem to face a forced choice between politically equal but relatively incompetent masses and relatively more competent but politically unequal elites. Any effort to consult the public faces a parallel dilemma. Who should be consulted? How should they be selected? In open town meetings, only the passionately engaged or mobilizedunlikely to be anything like a representative sampletend to show up. In scientific polling, random sampling assures representative samples, but their views, on most policy matters, are apt to be non- or minimal attitudes, closer, in truth, to dont-knows than to serious opinions. 7 Deliberative Polling helps solve these dilemmas, increasing both political equality and deliberation, both representativeness and real opinions. The basic idea is to assemble a random Political Information in Measures of Political Attitudes, ed. John P. Robinson, Phillip R. Shaver, and Lawrence S. Wrightsman (San Diego: Academic Press, 2000).
7

Philip Converse The Nature of Belief Systems in Mass Publics, pp. 206-61 and Attitudes and

Nonattitudes: Continuation of a Dialogue in Edward R. Tufte, ed., The Quantitative Analsys of Social Problems (Reading, Ma: Addison Wesley), pp, 168-89. Cf. Christopher H. Achen, Mass Political Attitudes and the Survey Response, American Political Science Review 69, 4 (December 1975): 1218-31 and Toward Theories of Data: The State of Political Methodology, in Ada W. Finifter, Political Science: The State of the Discipline (Washington: APSA, 1983); Jennifer Hill and Hanspeter Kriesi, An Extension and Test of Converses Black-and-White Model of Response Stability, American Political Science Review (June, 2001), 95: 397-413. For a recent overview, see George F. Bishop, The Illusion of Public Opinion: Fact and Artifact in American Public Opinion (Lanham, Md: Rowman and Littlefield, 2005).

sample to discuss a set of policy or electoral issues, having first sent them carefully balanced briefing materials and then giving them the chance of questioning panels of competing policy experts or policy makers. The small groups are randomly assigned and led by trained moderators. The participants answer the same questions before and after deliberating. 8 The random sampling serves equality and representativeness. The balanced information, moderated discussion, and questioning of balanced panels provide good conditions for deliberation. Of course the success of this initiative turns on the empirical evidence about what actually happens when citizens deliberate. Thus far, Deliberative Polling has been conducted mostly in Western democracies, including the US, Britain, Canada, Denmark, and Australia. Other projects have taken place in the newer democracies of Hungary and Bulgaria. 9 But these Eastern European countries have undergone transitions to democracy in the sense of party competition and its attendant conditions. In the case of China, the lack of party competition as well as limitations on the media and free expression provide a quite different context. In theory the Deliberative Polling research program provides a distinctive answer to the three questions with which we started. First, by using random sampling we provide an answer to
8

See James S. Fishkin, Democracy and Deliberation: New Directions for Democratic Reform

(New Haven, CT: Yale University Press, 1991) and The Voice of the People: Public Opinion and Democracy (New Haven: Yale University Press, 1997) for the idea and rationale; Robert C. Luskin, James S. Fishkin and Roger Jowell Considered Opinions: Deliberative Polling in Britain, British Journal of Political Science, 32: 455-487 for a detailed analysis; and James S. Fishkin and Robert C. Luskin, Experimenting with a Democratic Ideal: Deliberative Polling and Public Opinion, Acta Politica 40 (September, 2005): 284-298 for an overview.
9

See http://cdd.stanford.edu.

the who question that satisfies political equality. A random sample of the mass public, particularly if its representativeness is vindicated by who actually participates, can plausibly constitute a microcosm of the whole population. And it does so by providing everyone an equal chance of participation. Second, the process embodies good conditions for this random sample to come to considered judgments. The process employs balanced briefing materials, small group discussions with trained moderators, plenary sessions with competing experts who answer questions from the small groups from different perspectives, opportunities to reflect on the information and share competing points of view. The participants express their views in confidential questionnaires, insulated from social pressures. Deliberative Pollings answer to our third question, to what effect, has usually been less clear. The results may be broadcast and taken seriously by the media and even inform the debates among policy makers, yet fall well short of determining policy. That may trouble those wishing to restrict deliberative democracy to processes culminating in binding decisions. 10 We would dispute such a requirement, but, in any case, there have in fact been Deliberative Polls culminating in binding or all but binding decisions. Notably, Deliberative Polling led to the implementation of major investments by electric utility companies in Texas, instructed to follow

10

See Amy Gutmann and Dennis Thompson Why Deliberative Democracy? (Princeton:

Princeton University Press, 2004), p 5, although they define binding loosely: [The participants] intend their discussion to influence a decision the government will make or a process that will affect how future decisions are made (p. 5). By this criterion, many of the televised Deliberative Polls before elections or referenda, or advisory to government bodies would fall within the binding category.

the results by the states Public Utilities Commission. 11 It also was recently used by PASOK, the Greek Socialist party, to choose its candidate for Mayor of Marousi, (a large municipality in the Athens metropolitan area). A Deliberative Poll could be evaluated in various ways. It is usually several things at oncea social science investigation, a public policy consultation, a contribution to the media and public discussion. As a public policy consultation, we think some reasonable criteria for evaluation are: 1) The extent to which the sample is representative, 2) The extent to which there are significant changes in opinions, particularly about policy attitudes, 3) The extent to which the changes of opinion exemplify normatively desirable processes of deliberation. In particular, a) The extent to which the process avoids distortions from inequality, b) The extent to which the process avoids any predictable pattern of small group polarization,

11

See R.L. Lehr, W. Guild, D.L. Thomas, and B.G. Swezey Listening to Customers: How

Deliberative Polling Helped Build 1,000 MW of New Renewable Energy in Texas, National Renewable Energy Laboratory, June 2003. A much lengthier process designed to resemble Deliberative Polling, the Citizens Assembly in British Columbia, led to the selection of a referendum topic for electoral reform in 2004 (see Andr Blais, R. Kenneth Carty & Patrick Fournier, Do Citizen Assemblies Make Reasonable Choices? Paper presented at the Princeton Conference on Deliberative Democracy, Princeton University, Princeton, NJ, March 8-11, 2006.

c) The extent to which there is a development of public spirited preferences, d) The extent to which the participants learn, and e) The extent to which the learning drives the opinion change 4) The extent to which the post-deliberation opinions or changes of opinion influence public policy. Local Public Consultation in China In recent years, China has seen the development of consultative and deliberative institutions in the form of public hearings. The public has been invited to express its views on such local issues as the prices of water and electricity, park entry fees, the relocation of farmers, the development of historical sites, and even the possible relocation of the famous Beijing zoo. 12 In the mid- to late 1990s, some villages developed meetings in which village representatives discuss major decisions on village affairs. These practices have now spread to more urban communities. In the Shangcheng district of Hangzhou city, for example, a consensus conference or consultation meeting is held regularly once a month. 13 In one state-owned factory, a

12

Peng Zhongzhao, Xue Lan and Kan Ke, Public Hearing System in China, Beijing: Qinghua

University Press, 2004.

13

Baogang He, The Theory and Practice of Chinese Grassroots Governance: Five Models,

Japanese Journal of Political Science, vol. 4, no. 2, 2003, 293-314.

representative council of staff and workers deliberated for several months to decide the allocation of new departments to workers and managers. 14 Wenling City, the site of our Deliberative Poll, has been particularly active on this front. , holding numerous deliberative consultations called kentan (sincere heart-to-heart discussion). From 1996 to 2000, there were 1190 at the village level, 190 at the township level, and 150 in governmental organizations, schools and the business sector. 15 Some were connected to decision-making through the local Peoples Congress. Some similar practices have even sprouted at the national level. In 1996, the first national law on administrative punishment introduced an article on holding public hearings before punishments were taken. 16 The famous article 23 of the Law on Price passed by the Chinas National Congress in December 1997 specified that the price of public goods be decided through public hearings. This was followed by the Law on Legislature, passed in 2000, which required public hearings before passing any legal regulations or law. 17 More than 50 cities have now held legislative public hearings.

14

Jonathan Unger and Anita Chan, The Internal Politics of an Urban Chinese Work

Community: A Case Study of Employee Influence on Decision-making at a State-Owned Factory, The China Journal, No. 52, July, 2004, pp. 1-24.
15

See the official document, Democratic Sincerely Talk: The Innovation from Wenling, compiled

by the Department of Propaganda, Wenling city, 2003, p. 98.

16

Zhu Mang, Multiple Dimensions of Administrative Law, Beijing: Beijing University Press,

2004. Chapter one is devoted to the topic of public hearing on administrative punishment.
17

Wang Quansheng, A Study of Legislative Hearing, Beijing: Beijing University Press, 2003.

10

But public hearings have the same limitations in China as anywhere else. Above all, the participants are unrepresentative. Thus Cai Dingjian complains that public hearings do not really involve ordinary citizens and urges popularizing them, and Yang Zhongxin, the Director of the Price Bureau at Qinghuangdao city, argues that public hearings on prices often decide to increase prices because they are usually dominated by business interests. 18 The procedural requirements are also often vague, and there is often insufficient time for deliberation. In China, moreover, there is the additional danger of the dialogues being manipulated or the participants being selectively mobilized by officials. 19 The Deliberative Poll in Zegou Township, Wenling City How would a more ambitious model of deliberative democracy, like the Deliberative Poll, fare in a country like China? The design of Deliberative Polling, with its random sampling, extended deliberations, balanced briefing materials and expert panels, and clear aggregation rules for determining results, is intended to overcome the limitations of public hearings. Here we report on a Deliberative Poll sponsored by the local government in Zeguo Township of Wenling City, a county-level city with a vibrant private economy. Zeguo has an area of 63 square kilometers, of which the town centre is 6.5 square kilometers. It contains 89 villages, and 9 urban residential committees. The permanent local population is roughly 120,000,
18

Their speech at the international conference on public hearing in China, July 1, 2005. See Chen

Shengyong and Baogang He, eds., Development of Deliberative Democracy, Beijing: Chinas Social Sciences Press, 2006, p. 445, and p. 449.
19

Chen Shenyong and Baogang He, eds. Development of Deliberative Democracy, Beijing:

Chinas Social Sciences Press, 2006, 478 pages. The appendix includes a summary of the international conference on public hearings held July 2005.

11

and the floating (migrant) population another 120,000. The four major industries are shoes, water pumps, air compressors, and new building materials. The question the participants were asked to consider was which of a set of thirty possible infrastructure projects should be funded in the coming year. The projects included new bridges, roads, a school, and city gardens. In total, the projects would cost roughly RMB 137,000,000. Since only an estimated RMB 40,000,000 could be raised for urban planning, environmental and infrastructure construction, the local government had to prioritize. Depending on which projects were chosen, the available funding could cover only ten to twelve of the thirty possibilities. The idea, from the beginning, was to use Deliberative Polling as a way of democratizing local policy making. Thus the Zeguo town leadership madeand carried through onan explicit commitment to choose the projects the sample rated highest after deliberating. So this deliberation was effectively binding. A working committee composed of the Deputy Head of the Department of Propaganda in Wenling City, Dai Kangnian, Officer Chen Yiming, Party Secretary Jiang Zhaohua of Zeguo, and Deputy Party Secretary Wang Xiaoyu of Zeguo organized an expert committee that carried out a preliminary study of, wrote the feasibility reports for, and drafted briefing materials on the infrastructure projects. We helped local officials prepare the questionnaires and briefing materials, doing our best to ensuring that the latter were balanced and accessible and contained arguments for and against each project. The experiment began with an initial survey in March, 2005. A simple random sample of 275 Zeguo residents was drawn from a household registration list. The response rate (the proportion completing the initial interview) was a pollsters idea of heaven, the participation rate (the proportion of those who attended the deliberations) a Deliberative Pollsters idea of the

12

same. Of the 275, 269 completed the initial questionnaire, and 257 showed up on the day (April 9th, 2005). Of the latter, 235 also completed the final questionnaire). 20 Appendix A, comparing the 235 interviewees who attended the deliberations and completed the final questionnaire (participants) with the 34 who did not (nonparticipants), shows the participants to be an attitudinally representative subsample of the whole interview sample. On only one of the 30 projects (roughly the 5% expectable by chance) did the participants and nonparticipants enter the process with significantly different attitudes. To be sure, Appendix A also reveals some sizable and statistically significant differences between the and participants with respect to sociodemographic characteristics. Almost two-thirds of the participants but just over 80 percent of the nonparticipants were male. The participants averaged 47.5 years old, the nonparticipants 37.6 years old. Only about 20 percent of the participants but more than 50 percent of the nonparticipants had a high school education or beyond. More than 60 percent of the participants but only about 20 percent of the nonparticipants were farmers. Only 16.5 percent of the participants but 52.2 percent of the nonparticipants were entrepreneurs. But since there were only 34 nonparticipants, the participant sample still closely resemblesand in no wise differs significantly fromthe whole interview sample (which is in turn virtually the same as the entire sample). There was admittedly one respect in which the sample did manifestly differ from the population. There were too far too many men (although the participants were less unrepresentative in this respect than the nonparticipants). This resulted from a failure to implement one customary element of Deliberative (and other careful) Polling, namely random
20

A few participants were excluded from the analysis because they appeared to be cases in the

designated participant sent a family member or friend in his or her stead.

13

selection within the household. Instead, the household members exercised some discretion as to who would take the questionnaire. Subsequent Chinese Deliberative Polls (see below) have corrected this problem. The deliberation lasted one day. As in other Deliberative Polls, the design alternated small group and plenary sessions. In the small group sessions, the participants considered the advantages and disadvantages of each project and formulated key questions to put to the panels of competing experts in the plenary sessions. There were sixteen small groups, averaging about sixteen participants apiece. They were led by moderators (teachers selected from Zeguo High Schools) trained not to give any hint of their own opinions, to foster equal and civil discussion, and to facilitate the process of forming questions for the expert panels. At the end of the day, the participants completed an augmented version of the same questionnaire as they were given on first contact. Project Priorities The participants were asked to rate each of the 30 projects on a 10 point scale, with 0 being extremely unimportant, 10 being extremely important, and 5 being neither important nor unimportant. There was also a dont know (DK) option. Table 1 shows the mean ratings before and after deliberation. The scores are translated to a 0 to 1 scale. Twelve of the 30 projects showed statistically significant change at the .1 level (two-tailed). Generally speaking, the participants became more interested in sewage treatment and road construction that would affect their daily lives. All three sewage treatment projects received much higher support after deliberation. Some of these changes appear to reflect an increase in something like public spiritedness, about which we say more below. The average support for Wenchang Main Ave, a road crossing a number of villages increased by almost a full point. In

14

contrast, roads more specific to particular villages received diminished support. When it came to parks, a Peoples Park, for recreation, gained support, but Wenchang Park, a kind of town square that was touted as good for the citys image, lost support, as did commercial roads designed to connect factories with main roads. (Table 1 about here) For most ensuing analysis, we boil these projects down to ten broader categories, captured by mostly multi-project indices. Five are road-related. Industrial Roads includes roads in industrial areas and targeted to improve these industrial zones. Village Roads includes road constructions within specific villages. Main Roads includes roads traversing the whole town or important to most villages. Commercial Roads includes roads connecting factories with main roads. Wenchang Main Avenue is a single-item index. Two indices concern parks: Recreational Park, a single-item index, refers to a park for the entire township. Other Parks includes park constructions for specific villages. Sewage Treatment included four sewage treatment projects, all designed to serve the entire township. Township Image contains projects aimed at improving the townships appearance, for example by planting greenery and flowers. Cultural Heritage contains two projects (the reconstruction of Old Street and second stage construction of Wenchang Park) using traditional cultural architecture and designs. Appendix B lists the variables in each index and provides the inter-item correlations (for the two-item indices) and (Table 2 about here) Cronbachs alphas (for the multi-item indices). Consistent with Table 1, Table 2 shows the participants as coming to give greater priority to Wenchang Main Avenue and Sewage Treatment and lesser priority to Village Roads, Commercial Roads, Township Image, and Cultural Heritage. These six indices changed significantly at the .1 level or below (by a two-tailed test).

15

Public Spiritedness Discussions of deliberation and political participation have long speculated that forms of public consultation that involve shared discussion and decision about public issues will foster public spiritednessa greater support for policies of broad rather than narrow public interest. J.S. Mill, building on Tocquevilles account of town meetings and juries in America praised institutions that serve as schools of public spiritlocal decision making bodies where the interests of the whole community are discussed and individual citizens have some role in decisions. More recent writers like Jane Mansbridge have continued the speculation but have encountered difficulty finding clear empirical confirmation. 21 The Zeguo Deliberative Poll provides a good opportunity, in an unexpected context, to test these speculations. The projects varied a great deal in the proportion of the towns population they would benefit. A five-point scale was used to rate the extent to which each project would benefit the whole of Zegou Township. Projects benefiting only a small number of villages were rated as 1, projects benefiting a large number of villages as 5. The ratings are displayed in Appendix B. The correlation, across the 30 policy priorities, between this shared-benefit rating and the change in policy priority is .655. After deliberation, the participants priorities shifted toward projects benefiting the entire town. In that important sense, at least, they appear to have become more public-spirited.
21

See J.S. Mill Considerations on Representative Government (New York: Prometheus Books,

1991), especially chapters 1 and 8, pp. 78-9 and 171-3. See also Jane Mansbridge On the Idea that Participation Make Better Citizens in Stephen L. Elkin and Karol Edward Soltan, eds., Citizen Competence and Democratic Institutions (University Park, PA: Penn State University Press, 1999), 291-325.

16

Knowledge Gains The questionnaire contained four questions tapping the participants knowledge of the policy context in Zegou Township. These asked: (1) whether Zeguo Townships revenue had increased by 10.2 percent, 20.1 percent, 33.7 percent, or not at all from 2003 to 2004; (2) whether Zeguo Townships floating population is 50,000, 120,000, 200,000 and 300,000; (3) whether water pumps, shoes, plastic products, or air compressors are not a major product of the Township; (4) whether the Township has zero, one, two, five or seven parks. The correct answers were (1) 33.7 percent, (2) 120,000, (3) plastic products, and (4) two. The participants gained on all four items, significantly so on three of them. On average, the percentage answering correctly increased by eleven percent, which is highly significant. (Table 3 about here) Inequalities Some critics of deliberation, including Lynn Sanders and Iris Marion Young, have argued that the more privileged will dominate discussion and disproportionately influence the results, which should thus incline toward their views. 22 Such a skew would undermine the aspiration of deliberative democrats that everyones views get appropriate consideration on the merits. Critics of previous, less structured Chinese public consultations have noticed the same danger there. 23 One simple empirical approach to this question is to examine whether the sample as a whole tends to move toward the initial opinions of the more privileged or higher status
22 23

Lynn Sanders Against Deliberation and Iris Marion Young Intersecting Voices. For concerns about the inequalities in the current public hearing system see Chen Shengyong

and Baogang He, eds., Development of Deliberative Democracy, Beijing: Chinas Social Sciences Press, 2006, pp. 445 and 449.

17

participants. For purposes of this test, we take the more privileged to be the men, the more highly educated, and those in the most privileged occupations, in this setting the entrepreneurs and merchants. Table 4 shows that, far from moving toward the positions of the more privileged, the sample moved away from the time 1 position of the more highly educated on half of the indices, away from the time 1 position of the men on three-fifths of them, and away from the time 1 position of the entrepreneurs and merchants on four-fifths of them. At least in this setting, Deliberative Polling seems to create an environment in which inequalities in the broader society do not distort the deliberative process. (Table 4 about here) Polarization and Consensus in Small Groups Cass Sunstein has argued that there is a law of group polarization, according to which discussion predictably moves participants toward more extreme positions. A group beginning on one side of the midpoint will move further out in the same direction. This poses a normative challenge to deliberative democracy in implying that deliberation may change attitudes as a predictable artifact of group psychology rather than on the merits as the participants see them. Sunstein believes that there are two basic mechanisms by which discussions produce polarization in this sense. First, if the group begins on one side of the midpoint, the arguments voiced are likely to be balanced in favor of that side. Second, there is a social comparison effect. People will feel social pressure to agree with the perceived majority. Sunstein and various collaborators have confirmed these hypotheses with experiments with mock juries. 24

24

See Sunstein Deliberative Trouble and The Law of Group Polarization.

18

Yet this pattern has not held in previous Deliberative Polls. 25 The difference makes sense, for two reasons. First, the arguments to which the participants are exposed tend to be relatively balanced, thanks to balanced briefing materials, moderated small group discussions aimed at considering competing arguments, and balanced panels of competing experts. Second, there is less social pressure. The participants final opinions are solicited only in confidential questionnaires, and there is no common verdict to be reached. Does Deliberative Polling display its usual absence of polarization in China? Table 5 reports the movements toward or away from the midpoint for the ten priority indices in the 16 small groups. Overall, only 47.5% of the 160 group-issue combinations move away from the time 1 midpoint, about what one would expect by chance. In this Chinese context, too, the (Table 5 about here) Deliberative Poll belies the law of group polarization. Further experimentation may do more to establish the boundaries between polarizing and non-polarizing deliberations, but those in Deliberative Polling are clearly non-polarizing. One might worry if the members of a given small group always converged on a single position. A deliberation not steered toward consensus might nonetheless tend toward it on some issues but should not be normatively expected to do so on every issue. Much presumably depends on the degree to which the relevant interests and values are shared. In past Deliberative Polling projects, the within-group variance of opinion has not typically decreased too much more

25

See, for example, Robert C. Luskin, James S. Fishkin and Roger Jowell Considered

Opinions: Deliberative Polling in Britain, British Journal of Political Science, Vol 32, pp. 455487.

19

than half the time. 26 But what of these Chinese deliberations? The rightmost column of Table 5 shows that the variance within a larger than usual percentage (70.6%) of the small group-issue combinations do shrink. This is not large enough to be worrisome but does leave the question of why it is larger than in most previous Deliberative Polls. It may well be something about the nature of the issue. At a glance, the projects that particularly stood to benefit the whole community tend to be those for which the percentage of groups whose within-group variance decreases is largest. The mean percentage for the Sewage Treatment, Wen Chang Main Avenue, and Recreational Park indices is 89.6%. For the remaining seven indices, it is only 62.5%. At the level of the 30 individual projects, the correlation between the shared-benefit scale introduced above and the percentage of groups showing a decrease in within-group variance is .429. At the level of the ten indices, the correlation is .142. In this light, the tendency toward increased agreement, concentrated as it is on projects benefiting the whole community, would appear to be another aspect of the increase in public-spiritedness. Information and Opinion Change A simple model can estimate the extent to which the participants who emerged with the most knowledge were the ones that changed most. 27 The model is P2 - P1 = 0 + 1K2 + 2(P1 - G1) + u, where P1 and P2 are the participants positions at T1 and T2 (before and after deliberation), K2 is his or her knowledge at T2, G1 is the mean position of the participants small group (disregarding

26 27

Luskin et al, Considered Opinions p. 477. Luskin et al, Considered Opinionspp. 480-83.

20

the participant him- or herself) at T1; 0, 1, and 2 are the parameters; and u is a disturbance. 28 We use T2 knowledge rather than the T1-T2 knowledge gain, because the combination of item sampling bias (toward easy questions), ceiling effects (a respondent answering all questions correctly at T1 cannot show any increase), and the the rich getting richer (those knowing the most at T1 being those who actually learn the most) implies that true knowledge gain is better proxied by observed T2 knowledge than by observed knowledge gain. 29 Normatively, we should want the T2 knowledge coefficient 1 to have the same sign as the mean opinion change P2 - P1, meaning that those who emerge knowing the most are disproportionately responsible for the overall change. Theoretically, we should also expect though not necessarily want the small group coefficient 2 to be negative, meaning that participants are narrowing the gap between their own and their small groups T1 position. Table 6 reports the ordinary least squares estimates for the six project indices showing significant change. The signs of the estimated coefficients are all as expectedthose for the small group variable P1 - G1 always negative, those for T2 knowledge always sharing the sign of the overall opinion change. All six of those for the small group variable are significant (at the .01 level), as are four of those for T2 knowledge (three at the .01 level and one at the .1 level). At least on the surface, the participants do seem to be narrowing the gaps between their own and their small groups mean T1 opinion. It should be noted, however, that if the P1 - G1 variable is split apart, and the model re-estimated with K2, P1, and G1 as separate regressors, almost all of
28 29

All the variables are implicitly subscripted for the ith participant and jth project index. See Luskin et al., Considered Opinions, pp. 480-81, and Robert C. Luskin. True versus

Measured Information Gain, manuscript, Department of Government, University of Texas, Austin, TX 78712, USA.

21

(P1 - G1)s effect turns out to belong to P1, whose negative coefficient can be interpreted as mere regression toward the mean. 30 How much to make of the P1 - G1 coefficient estimates is therefore unclear. What is clear, from the K2 effects, is that the changes in the priorities awarded these projects are, in large measure, learning-driven. The more information one emerges with, the more one contributes to the overall opinion change. (Table 6 about here) Conclusion The public criteria for policy implementation in China are that it be scientific, democratic and legal. The Zeguo Deliberative Poll was scientific, in using social science to consult the public; democratic in offering the voice of a random sample, not just the party cadres; and legal in submitting the results to the local Peoples Congress, which approved them overwhelmingly, before they were implemented. More importantly from our perspective, the Zeguo Deliberative Poll seems to have done very well on all four of the criteria proposed above. First, the sample was highly representative. The selection was random, except within the household, which led to the one notable bias: too many men. This defect has already been remedied in a subsequent (2006) Deliberative Poll in Zeguo. Second, deliberation brought opinion change. Third, the opinion change exhibited several normatively desirable properties. There was no tendency to change in the direction of the opinions held by higher status or more privileged participants. There was no consistent pattern of polarization. There was an increase in public-spiritedness, in the sense that the participants grew more interested in projects benefiting the broader community rather than just their own villages. The participants became more informed. And the opinion changes and information gains were
30

Just as in Luskin et al., Considered Opinions. Results available on request.

22

related. Those who emerged knowing the most were disproportionately responsible for the overall changes of opinion. Lastly, the results were a decisive input into the policy process. All twelve projects selected are in the process of being built. Ironically, some of the legacies of authoritarian rule made it easier to satisfy some of these criteria. The expectation of participation for public purposes made it easier to recruit the sample. Though paid only a modest fee for their days participation, 31 they also knew that they were expected to come. In addition, the autonomy of local party officials to implement policy made it easy for them to deliver on a promise to implement the results. The results did surprise them. Jiang Zhaohua, the Zeguo Town Party Secretary, did not expect the high ratings for sewage treatment and other environmental projects, nor the low ratings for image or road projects. Eight of ten environmental projects but only one (Wenchang Main Avenue) of seventeen road-related projects wound up in the top ten. 32 In general, he was surprised at the difference between the local leaderships perception of public opinion and actual public opinion. Yet the local leadership was pleased with the processin the first place for its deliberative properties and, in the second place, for providing a way of responding to deliberative preferences. Ye Qiquan, the head of Zeguo town, initially unenthused about Deliberative Polling, saw the participants as increasing their understanding of the projects, thinking about

31 32

They were paid 50 Chinese Yuan each, equivalent to around US $ 6. Jiang Zaohua and He Baogang, "Deliberative Democracy: The Participatory Decision-making

Mechanism," Chen Shengyong and Baogang He, eds., Development of Deliberative Democracy, Beijing: Chinas Social Sciences Press, 2006, pp. 227-228.

23

which to prioritize, and acquiring more of a community-wide perspective in the process. 33 Jiang Zhaohua observed, Although I gave up some final decision-making power, we gain more power back because the process has increased the legitimacy for the choice of priority projects and created public transparency in the public policy decision-making process. Public policy is therefore more easily implemented. At least in the current Chinese context, he was undoubtedly right. A nearby town that did not consult the public about giving land to chemical plants faced protests, even riots as villagers blocked roads. But Zeguo benefited from local support and a sense that the government was responding to the public needs voiced by the people. 34 It is a measure of the Zeguo Deliberative Polls success, and of Deliberative Pollings promise in China, that this first Zeguo Deliberative Poll was followed by a second the following year, on March 20, 2006, to help select that years infrastructure projects. Again a scientific sample was gathered, became more informed, and deliberated on the merits of the projects. Again the results demonstrated substantial concern for the environment. And a further policy consequence was that Jiang Zhaohua appointed an official to take charge of environmental affairs and allocated about one million further Chinese Yuan for environmental projects. 35
33 34

Personal communication to the authors. Howard W. French Chinas New Frontiers: Tests of Democracy and Dissent, The New York

Times, June 19, 2005.

35

These two Zeguo Deliberative Polls also inspired one other, more specialized effort, also in

Zeguo. In July, 2006, a public official aware of the Zeguo township Deliberative Polls organized a Deliberative Poll among the workers in a factory, with the reducing turnover by better addressing the needs of the workers.

24

Whether widespread Deliberative Polling would contribute to democratization in China is an open question. 36 It does nothing, directly, for party competition. But it can promote the notions that government can be responsive to public needs and that citizens can voice their views in a context of equality and mutual respect. It could contribute to democratic development over the long term by educating participants and observers in the ways of democratic citizenship and giving them a sense of empowerment. Or it could retard democratic development by contributing to the legitimacy of existing institutional structures. These are complex and uncertain issues. In the meantime, however, this project helps demonstrate that deliberative democracy can be realized, at least for a microcosm, even outside established democracies.

36

Suzanne Ogden notes the importance of deliberation in the Chinese political system as a

means of reaching consensus and this deliberation could prove to be an important building block for democratization. Suzanne Ogden Inklings of Democracy in China, Harvard University Press, 2002, p. 257. Professor Lin Shangli, the dean of social sciences at Fudan University argues for deliberation-led democratization. See Lin Shangli, Deliberative Politics: A Reflection on the Democratic Development of China. Academic Monthly (Shanghai) 4: 1925, 2003. A new edited book focuses on this issue. See Ethan Leib and Baogang He, eds., The Search for Deliberative Democracy in China, New York: Palgrave, 2006.

25

Table 1: Project Priorities Question Wenchang Main Ave First Stage of Muchang Main Road Bridge Fuxin Road (east end) Dongcheng Road (first gate) Dongcheng Road (second stage) Shuangchen Road (first gate) Shuangchen Road (second stage) Tngquao Road Reconstruction for Donghe road Donghe main Ave Xicheng Road (first stage) Zeguo main Ave (second stage) Zeguo main Ave (third stage) Air compressor industrial zone matching environmental constructions Auxiliary environmental construction for Muyu, Lianshu, & Shuichang industrial zones Chengqu subroad rebuild Guojialing hill side reconstruction Wenchang park (first stage) Wenchang park (second stage) Citizen park (first stage) Urban environmental constructions Danyan hill park Muyu hill park Urban & countryside environmental projects Demonstrative street Old street reconstruction Sewage Treatment Plan, Muyu Sewage Treatment Plan, Danyan Sewage Treatment (earlier stage) entire town N 160 116 111 96 90 99 110 96 86 93 101 108 110 93 81 114 97 86 109 98 109 118 134 118 134 114 111 133 145 167 T1 .825 .688 .742 .578 .543 .561 .697 .600 .502 .714 .563 .626 .583 .467 .563 .667 .568 .560 .593 .518 .696 .755 .761 .721 .864 .675 .637 .729 .753 .892 T2 .924 .554 .706 .505 .510 .459 .612 .466 .473 .583 .533 .630 .597 .459 .506 .689 .520 .595 .505 .350 .744 .731 .723 .704 .924 .649 .576 .886 .914 .971 T2-T1 .098 -.134 -.036 -.072 -.033 -.102 -.085 -.134 -.029 -.131 -.031 .004 .015 -.008 -.057 .023 -.048 .035 -.088 -.168 .048 -.024 -.038 -.017 .060 -.025 -.061 .157 .161 .080 s.e. .023 .054 .063 .042 .045 .042 .040 .045 .045 .044 .045 .041 .039 .047 .047 .033 .043 .044 .044 .045 .034 .038 .037 .032 .027 .045 .045 .032 .030 .022 p .000 .015 .571 .084 .458 .018 .035 .004 .519 .003 .501 .928 .709 .872 .226 .496 .268 .429 .046 .000 .158 .530 .305 .593 .026 .572 .178 .000 .000 .001

26

Table 2: Project Indices N 153 158 173 121 160 174 109 176 136 194 T1 .623 .597 .624 .642 .825 .714 .696 .663 .590 .829 T2 .610 .538 .604 .562 .924 .684 .744 .618 .491 .921 T2-T1 -.013 -.058 -.020 -.080 .099 - .030 .048 -.045 -.099 .092 s.e. .029 .032 .025 .032 .023 .027 .034 .025 .035 .017 p .656 .066 .433 .015 .000 .270 .158 .076 .005 .000

Industrial Roads Village Roads Main Roads Commercial Roads Wenchang Main Ave Other Parks Recreational Park Township Image Cultural Heritage Sewage Treatment

27

Table 3: Information Gains T1 .204 .391 .421 .230 .312 T2 .315 .528 .494 .362 .424 T2-T1 .111 .136 .072 .132 .112 s.e. .037 .039 .038 .036 .026 p .002 .001 .028 .000 .000

Revenue increase in Zeguo, 2003-2004 Floating population in Zeguo Not a major product of Zeguo Number of parks in Zeguo Summary Index Note: n = 235, p-values one-tailed.

28

Table 4: Inequality Movements

Indices Industrial Roads Village Roads Main Roads Commercial Roads Wenchang Main Ave Other Parks Recreational Park Township Image Cultural Heritage Sewage Treatment Overall Movement Toward Away

N 153 158 173 121 160 174 109 176 136 194

T1 .623 .597 .624 .642 .825 .714 .696 .663 .590 .829

T2 .610 .538 .604 .562 .924 .684 .744 .618 .481 .921

T2-T1 -.130 -.500 -.199 -.798 .988 -.295 .477 -.464 -.993 ..092

p .656 .066 .433 .015 .000 .270 .158 .076 .005 .000

Male T1 .606 .618 .625 .651 .805 .698 .668 .650 .619 .830

High Education T1 .618 .661 .609 .650 .841 .732 .722 .664 .638 .873

Entrepreneur/ Merchant T1 .632 .601 .632 .657 .760 .728 .698 .645 .642 .835

40.0% 60.0%

50.0% 50.0%

20.0% 80.0%

29

Table 5: Small Group Mechanisms Percentage of groups moving away from the T1 midpoint 43.8 37.5 50.0 50.0 43.8 43.8 43.8 0.0 87.5 75.0 47.5 Percentage of groups whose within-group variance decreases 75.0 62.5 68.8 43.8 62.5 50.0 75.0 100.0 87.5 81.3 70.6

Indices Industrial Roads Village Roads Main Roads Commercial Roads Other Parks Township Image Cultural Heritage Sewage Treatment Wen Chang Main Ave Recreational Park Overall

30

Table 6: Attitude Change as a Function of T2 Information, T1 Small Group Mean and T1 Attitude Explanatory Variable Intercept T2 Information Rs distance from T1 group mean adjusted R2 F Probability n Village Roads (-) 1.095*** (.376) -3.045*** (.627) -.707*** (.072) .446 63.30 0.000 156 Commerc ial Roads (-) 1.029** (.520) -.717 (.805) -.591*** (.092) .264 22.15 0.000 119 Township Image (-) 1.103*** (.322) -1.412*** (.525) -.642*** (.071) .338 45.14 0.000 172 Cultural Heritage (-) 1.650*** (.461) -2.744*** (.734) -.732*** (.073) .452 55.89 0.000 134 Sewage Treatment (+) 2.076*** (.187) .356 (.300) -.749*** (.045) .595 141.26 0.000 192 Wen Chang Main Ave (+) 2.758*** (.317) .698* (.463) -.786*** (.061) .525 87.84 0.000 158

Note: Cell entries are coefficient estimates with estimated standard errors in parentheses. The parenthetical signs in the column headings indicate the direction of net change for the sample as a whole and thus the expected sign of the information coefficient. t2 information is the mean of the four information items. The group mean variables are calculated on the other group members, excluding the respondent. * Significant at the 0.10 level ** Significant at the 0.05 level *** Significant at the 0.01 level

31

Appendix A: Representativeness 1. Demographics Variable Entire Sample (n = 269) Participants (n = 235) Non-Participants (n = 34)

Gender Male* Age** Martial Status Married Education High School or Beyond** Occupation Farmer** Worker Entrepreneur (business owner)** Merchant Teacher Public Servant Other

70.1 42.6 94.0 24.3 60.0 3.9 21.0 8.3 2.0 1.5 3.4

66.2 47.5 92.9 20.8 62.8 3.7 16.5 7.4 1.6 1.6 3.7

80.8 37.6 92.0 51.8 21.7 4.3 52.2 13.0 4.3 4.3 0.0

Note: Entries are percentages except for age, which is in years. *Difference between participants and nonparticipants significant at the .10 level **Difference between participants and nonparticipants significant at the .01 level There are no statistically significant differences between the participants and the whole sample. 2. Attitudes Question Wenchang Main Ave First Stage of Muchang Main Road Bridge Fuxin Road (east end) Dongcheng Road (first gate) Dongcheng Road (second stage) Shuangchen Road (first gate) Shuangchen Road (second stage) Tngquao Road Reconstruction for Donghe road Donghe main Ave Xicheng Road (first stage) Zeguo main Ave (second stage) Zeguo main Ave (third stage) Air compressor industrial zone matching environmental constructions Auxiliary environmental construction for Muyu, Lianshu, & Shuichang industrial zones Participants at T1 ..807 .634 .727 .550 .546 .546 .685 .576 .502 .715 .586 .627 .592 .460 .567 .664 Non-participants at T1 .796 .524 .682 .535 .647 .757 .668 .610 .606 .608 .667 .600 .583 .445 .561 .738 P-NP .011 .110 .045 .015 -.102 -.211 .017 -.034 -.104 .107 -.081 .027 .009 .015 .006 -.074 S.E. .058 .078 .071 .081 .083 .079 .075 .082 .095 .075 .084 .090 .086 .089 .090 .075 p .849 .157 .526 .858 .220 .009 .823 .682 .274 .157 .341 .768 .920 .866 .944 .327

32

Chengqu subroad rebuild Guojialing hill side reconstruction Wenchang park (first stage) Wenchang park (second stage) Citizen park (first stage) Urban environmental constructions Danyan hill park Muyu hill park Urban & countryside environmental projects Demonstrative street Old street reconstruction Sewage Treatment Plan, Muyu Sewage Treatment Plan, Danyan Sewage Treatment (earlier stage) entire town Note: Entries are means; p-values are two-tailed.

.549 .521 .602 .514 .692 .753 .747 .708 .862 .682 .653 .743 .751 .876

.547 .557 .643 .570 .673 .777 .781 .754 .919 .660 .691 .771 .800 .904

.002 -.036 -.041 -.056 .020 -.024 -.034 -.046 -.057 .022 -.037 -.029 -.049 -.028

.087 .088 .082 .082 .077 .064 .076 .072 .056 .088 .084 .079 .077 .060

.986 .683 .617 .497 .799 .713 .653 .525 .304 .799 .657 .715 .528 .645

33

Appendix B: Policy Indices Industrial Roads ( = 0.66) Tngquao Road (1) Air compressor industrial zone matching environmental constructions (1) Auxiliary environmental construction for Muyu, Lianshu, and Shuichang industrial zones (1) Village Roads ( = 0.64) First Stage of Muchang Main Road (3) Dongcheng Road (first gate) (2) Dongcheng Road (second stage) (1) Main Roads ( = 0.73) Reconstruction for Donghe road (1) Donghe main Ave (2) Xicheng Road (first stage) (2) Zeguo main Ave (second stage) (2) Zeguo main Ave (third stage) (2) Chengqu subroad rebuild (1) Commercial Roads (r = .54) Shuangchen Road (first gate) (2) Shuangchen Road (second stage) (1) Other Parks ( = .60) Wenchang park (first stage) (3) Danyan hill park (3) Muyu hill park (3) Township Image ( = 0.60)) Bridge (2) Fuxin Road (east end) (2) Wenchang park (second stage) (2) Urban environmental constructions (4) Cultural Heritage (r = 0.37) Wenchang park (second stage) (4) Old street reconstruction (2) Sewage Treatment ( = .67) Urban & countryside environmental projects (5) Sewage Treatment Plan, Muyu (5) 34

Sewage Treatment Plan, Danyan (5) Sewage Treatment (earlier stage) entire town (5) Wenchang Main Ave Wenchang Main Ave (4) Recreational Park Citizen park (first stage) (4) NOTE: Individual projects ratings on the shared-benefits scale are given in parentheses, following the projects name.

35