Deliberative Planning for Disaster Recovery: Re-membering New Orleans

A strong turnout of a broad cross-section of the New Orleans population lent legitimacy to a unique public conversation about post-Katrina recovery priorities. Designed and conducted by AmericaSpeaks, Community Congress II brought together over 2500 New Orleanians, linked electronically across five different cities plus smaller satellite sites in 15 diaspora communities across the country. Table observations of 48 deliberative rounds, along with exit interviews, post-event focus groups with participants, and stakeholder interviews are used to address three central questions about the social processes of deliberation: What are the patterns of interaction at the tables across race and gender; how do the participants interpret the meaning of the event for themselves; and what were the outcomes in terms of legitimacy, influence, and social trust? The article contextualizes the event in the unfolding story of the recovery process and culls out the lessons learned for deliberative democracy. Community Congress II demonstrates that inclusive public deliberation does more than provide reasoned public input into difficult policy decisions. It does more than legitimate new public initiatives. It can foster social trust and social healing across the divides of race, culture, and wealth. But the benefits will be sustained only if they are reinforced by an institutional infrastructure of civic engagement.


Introduction
New Orleans presents a challenging arena for deliberative planning. In a city with deep divides across race, wealth, and education, social trust is low. In a city known for corruption among public officials, mistrust of government runs high. Katrina conspiracy theories about the breaking levees and the stranded residents thinly veil a collective pain of betrayal and abandonment. In this setting we examine the first major attempt at wide scale public deliberation about disaster recovery.
A strong turnout of a broad cross-section of the New Orleans population, not only in New Orleans itself but in four diaspora cities, lent legitimacy to a unique public conversation about recovery priorities. Community Congress II, the deliberative forum for public input on city-wide recovery priorities, was held on December 2, 2006, as part of the official Unified New Orleans Plan (UNOP) process. Designed and conducted by AmericaSpeaks, it brought together over 2500 New Orleanians, linked electronically across five different cities plus smaller satellite sites in 15 diaspora communities across the country. Community Congress II became the pivotal point in creating public support for the city-wide recovery plan, after two failed attempts by the mayor and the city council. Our research team took an intimate look at the gathering in New Orleans to draw out the lessons for deliberative democracy.

Methodology and Research Questions
What intrigued me most about doing this research was the opportunity to observe the black box of civic engagement: the actual deliberative process at the tables. 1 Our research team was given direct access for continuous observation of 16 tables at the New Orleans site. 2 We were able to document patterns of interaction in the three discussion rounds held simultaneously at the 16 tables, yielding data on 48 deliberative discussions. We also conducted pre-and/or post-event interviews with 17 stakeholder representatives from the community, private sector, city government, and the UNOP planning team. We conducted 28 exit surveys with attendees and held three post-event focus groups with six to eight attendees in each. In addition we had access to the keypad data collected by AmericaSpeaks on demographics and voting. 3 1 Stromer- Galley (2007) and Ryfe (2005) point out a large gap in the public deliberation literature about the actual process of intra-group interaction. 2 The tables were selected arbitrarily by the individual table observers, mostly before the tables filled up with attendees. All 16 tables had a mix of African Americans and Caucasians. 3 We also gathered data at the Dallas and Houston meetings, including four table observations in Dallas and one in Houston. This article focuses on the New Orleans experience, where I had been involved in various neighborhood recovery planning efforts since late 2005.
Three central questions guided our research: 4 What are the patterns of interaction at the tables? Which demographic groups speak the most and which the least? What characterizes the conversations? How do the participants interpret the meaning of the event and their own participation? What were the outcomes? We looked at legitimacy, influence, and social trust/social healing. In addition, we contextualized the event in the larger unfolding story of the recovery process and culled out the lessons learned for deliberative democracy.

The Context
Shortly after hurricane Katrina left its devastating flood waters covering most of New Orleans on August 29, 2005, Mayor Nagin appointed a high profile commission called Bring New Orleans Back (BNOB) to prepare a recovery and rebuilding plan. 5 The land use committee, headed by an influential New Orleans businessman, Joseph Canizero, brought in a team of national experts from the Urban Land Institute to create a technically sound land use plan for New Orleans' recovery. The ULI plan, unveiled at a well attended conference in New Orleans in November, sparked an intense public outcry from a traumatized public: -We don't want outsiders or experts to tell us what to do! And we don't trust wealthy businessmen!‖ The final BNOB plan presented in January provoked even more vociferous opposition. Green zones dotted the city map showing which low-lying neighborhoods would not be rebuilt in order to ‗shrink the footprint' to higher ground. -You can't tell us where we can and can't rebuild. These are our homes!‖ BNOB responded by giving former residents 4 months to prove their devastated neighborhoods were viable. The announcement, which galvanized an intense effort of grassroots neighborhood planning across the city, further alienated the public. The mayor backed away from the BNOB plan.
As the mayor's planning effort collapsed, the City Council launched its own effort, known as the Lambert Plan. With a sole-source contract to a Florida consulting firm known to the Council, the Lambert group was brought in to work with the devastated neighborhoods in designing their recovery plans. While the closed contracting process was criticized, the Lambert team was successful in bringing in local architects and planners and working closely with neighborhood residents who were pleased to get the technical assistance. The visioning process excited the participants and created a sense of ownership of the plans (although non-participants criticized the plans as being unrealistic and misleading). But the Council did not have major stakeholders on board behind the Lambert Plan-neither the mayor, the City Planning Commission, nor the private investment sector. Moreover, because the Lambert Plan excluded the less impacted neighborhoods and districts, it would not add up to a city-3 wide recovery plan and therefore would not garner Louisiana Recovery Authority (LRA) approval for Federal funding.
Even as Lambert continued to work with the flooded neighborhoods, LRA quietly began to pull together the local political and institutional actors for a Unified New Orleans Plan, or UNOP as it would be called. Rockefeller Foundation offered $3.5 million to back a comprehensive New Orleans planning effort if the various players could be brought together to support it. A highly respected and politically neutral local foundation, the Greater New Orleans Foundation (GNOF), agreed to spearhead UNOP through an oversight board with fiduciary responsibility (New Orleans Support Foundation) and a citizens advisory board (Community Support Organization). GNOF contributed another $1 million to the effort and so did the Bush-Clinton Fund. In April 2006 the LRA hired Concordia, a local architectural and planning firm with a national reputation and experienced in public participation, to carry out the UNOP charge.
The UNOP process got off to a rocky start with fears from the flooded neighborhoods that their planning efforts would be undone, a full page ad in the newspaper by Lambert attacking UNOP, a chaotic introductory meeting with the neighborhoods at the end of July 2006, and delays in signing the memorandum of agreement with City Council, the Mayor, and the City Planning Commission until the end of August, 2006.
The UNOP design provided two levels for public participation: at the district level and at the city-wide level. The UNOP staff gathered and made available all previous neighborhood plans and required the district planning teams, who were selected with neighborhood input after a national RFQ, to incorporate them in their planning efforts. 6 The district level planning took off in each of the 13 districts with a series of four public planning meetings, interspersed with more numerous district stakeholder meetings in each district, with active citizen participation in all 13 districts. In some cases the planners built upon the community organizational infrastructure already in place for recovery planning. In others they created new steering committees and stakeholder groups. These meetings, which afforded local residents input into selecting key recovery projects for their neighborhoods and districts and a chance to discuss trade-offs, became the place of much detailed debate around the concrete implications of some of the cross-cutting city-wide issues. 7 The Lower Ninth Ward, for example, held four district meetings averaging 100 participants each, and 6 6 A controversy around the Lower Ninth Ward planning team, which was replaced early in the UNOP process, resulted in the development of an alternative plan by some of the original team members, according to Cornell planner Ken Reardon. The People's Plan was accepted for integration into the UNOP plan by City Council in February, 2007, and  The other arena for public participation was the city-wide planning process, with three city-wide congresses envisioned. 8 The New Orleans Support Foundation voted in early September, 2006, to have AmericaSpeaks facilitate the first congress and take full charge of the other two, including raising the funds to include the diaspora population in decision-making at multiple sites. 9 The hastily organized and poorly attended Community Congress I was held in New Orleans in late October, 2006, to update the public on the Katrina planning team's baseline studies and gather public input on a slightly re-phrased version of the volatile footprint issue. When the 300 attendees were asked -who's here?‖ the keypad responses showed a mostly white and upper income audience. Public and media criticism over the lack of representation (and the revival of the footprint issue) was vociferous. If the Unified Plan was to have credibility with the public, Community Congress II would have to turn it around. AmericaSpeaks raised the estimated $2.4 million needed for the multi-site event. 10 The Design of Community Congress II Outreach Much of the budget for CCII was earmarked for an extraordinary outreach campaign in five diaspora cities in three states plus 16 satellite locations across the country. The public information campaign was only part of it: Under a tight timeframe AmericaSpeaks hired subcontractors to cover both web messaging (websites, blogs, and email blasts) and traditional media messaging (radio, television, newspapers, magazines, newsletters, and mail-outs) in the five main cities. To assure turn-out from low income communities, AmericaSpeaks also developed a network of paid grassroots organizers to work with community-based and faith-based organizations, neighborhood organizations, and cultural organizations. The strategy was to deliver -clear messaging by trusted messengers‖ (key informant interview). 11 Canvassing 8 Getting agreement from GNOF and the city-wide planning team on the need for such widespread community input and the highly interactive method of AmericaSpeaks was not easy, according to a key informant from Concordia. 9 According to key informants, the idea of a multi-site interactive planning event with AmericaSpeaks had its inception in November 2005. Contracted by LRA to facilitate an interactive symposium in New Orleans on state-wide recovery principals organized by the American Institute of Architects, AmericaSpeaks was approached by the Committee for Better New Orleans with the idea of a multi-site meeting to include the diaspora population. Early efforts to pursue this idea fell in the shadow of political campaigns and the beginning of the City Council-sponsored Lambert planning process. 10 Rockefeller Foundation contributed $500,000 to CCII. 11 Organizers were paid in part by number of people registered, according to a representative from AmericaSpeaks, and registrations substantially outstripped actual attendance at the New Orleans site. and flyer activity, public signage, small group briefings, and resource centers in public libraries also targeted low income neighborhoods, such as the Lower Ninth Ward. The resulting outreach capability is being maintained by AmericaSpeaks for future outreach efforts, according to an AS representative.
On the day of the event, buses brought registered participants from many neighborhoods and apartment complexes in all five major sites to and from the event locations. Child care was available at each site, and breakfast and lunch were served.

The Design of the Day
CCII followed the protocol of the AmericaSpeaks 21 st Century Town Meeting, 12 which had gained visibility in 2002 at the -Listening to the City‖ event in New York about the World Trade Center site with 5,000 participants. In the AmericaSpeaks protocol the participants sit at tables of 8 to 10, each table with its own facilitator, to discuss pre-designed questions in 15 to 30 minute rounds. The discussion of each question is focused around three options listed in a discussion guide along with pros and cons for each option. Participants may also propose new options or combinations of options. Their ideas are collected on a laptop at each table and fed to a central ‗theme team' which synthesizes and reports back the common ideas of each round within minutes. The participants then choose their top three to five priorities from the expanded list of options, voting with individually assigned electronic keypads. The results are tabulated and reported back to the plenary nearly instantly. The New Orleans Community Congress II added a new dimension: simultaneous interconnected events in the four diaspora cities and 16 satellite locations across the country.
Community Congress II was a multifaceted production, carefully orchestrated, choreographed, and timed, designed to engage the participants at the level of head and heart. Thus it was much more than a deliberative event aimed at rational discourse around informed choice-making. Interwoven among the informational presentations and discussion cycles were motivational speeches from local civic and political leaders, a conversation about what we most treasure about New Orleans, a rap performance (-New Orleans is not just where I live--it's who I am‖), a poem based on participants' words, the presentation of a painting created in the room that day, live interviews of diverse participants at each of the five main sites projected onto big screens across all the sites, panoramic shots of the crowds waving and clapping at each location, and of course music-the music that defines and connects New Orleanians. An elaborate electronic multi-media infrastructure undergirded the live day-long interactive production. 13 12 The term is trademarked by AmericaSpeaks. 13 AmericaSpeaks itself uses the terminology of production: executive producer, associate producer, etc. An excellent short video montage of the New Orleans event that captures the multiple dimensions is available at www.youtube.com/watch?v=jA8xIVZAok8.

Framing the Questions
During the last week before CCII the policy questions to be deliberated were framed and a short discussion guide was written. 14 The AmericaSpeaks staff simplified the questions and options that had been developed by the city-wide planning team. A major concern of the planning team was to sidestep or reframe the volatile ‗footprint' issue (no-build zones in the most flood-prone areas) that had been the demise of the mayor's recovery plan. This concern was addressed in part by making it clear that the UNOP plan would not have a land use map nor attempt to define a vision for a new New Orleans; rather, it would be a strategic policy plan oriented toward identifying and prioritizing investments, as required by LRA to meet Federal guidelines. For this UNOP needed public input and support for major policy choices.
The footprint issue was obliquely embedded in a series of interrelated policy questions such as flood risk reduction and neighborhood stability. Controversy was to be reduced further by asking participants not to choose one policy option (which would create winners and losers), but to express their level of agreement or disagreement on each option and to propose new options from the tables.
The following six questions were put forward for public deliberation: To reduce the risk of property damage from flooding, the City should: a. provide information and let individuals decide how and where to rebuild b. provide financial incentives to individuals to reduce their risk c. create and enforce building standards and programs to reduce risk To rebuild infrastructure, the City should: a. distribute available funds evenly across city b. concentrate available funds in areas with greatest need c. raise additional funds to repair and improve all infrastructure To rebuild stable neighborhoods, the City should: a. provide information for homeowners to decide how and where to rebuild b. provide financial incentives for homeowners to rebuild near each other c. set and enforce standards for homeowners to rebuild To create rental and affordable housing, the City should a. rely on market forces and existing programs b. provide financial incentives to developers c. fund low and moderate income public housing To rebuild schools, hospitals, and clinics, the City should a. distribute facilities evenly across city b. reopen and rebuild based on repopulation and recovery rates c. reduce costs through multipurpose school/community facilities 15 14 The discussion guide offered a paragraph of background on each question, three to four options, and a brief listing of pros and cons for each option. The planners provided additional background information through a power point presentation before each discussion cycle. 15 A key informant from Concordia explained that the specificity of the communities-in-schools option reflected the fact that it was a high priority for their firm.
To rebuild police, fire, and criminal justice services, the City should a. distribute facilities evenly across city b. reopen and rebuild based on greatest needs c. reduce costs with multipurpose facilities The six questions would be discussed over the day in three deliberative rounds of 20 to 30 minutes each. At the end of each round, participants would use the keypads to register their level of support for each of the three official options. Then they would be shown a list of nine options that included at least six that the theme team had collected and synthesized from the tables. Participants would choose their top three priorities from that expanded list. At the end of the day, they would have the chance to select the five options from the entire day that they felt were the most important to rebuilding the city.

Process Findings
Representation: Who Came to CCII?
The racial and ethnic diversity at the New Orleans convention center on December 2, 2006, was visible at a glance, a welcome contrast to CCI held just weeks before. The confirmation came at mid-morning when nearly 1000 of the more than 1500 participants at the five major sites entered on their keypads the individual answers to the collective question -Who's here?‖ Together they represented a remarkable microcosm of the racial and ethnic composition of pre-Katrina New Orleans. 16 Demonstrating the fruits of an effective outreach program to low income residents, one in every four respondents indicated family earnings below $20,000, a sizeable proportion even though it was ten percentage points below the pre-Katrina proportion. Renters were the most underrepresented group. 17 The gathering in New Orleans itself mirrored the stark reality of the demographics of diaspora: a much larger representation of whites, homeowners, and high income earners than in the diaspora sites, and a much smaller representation of low income earners, African Americans, and renters. It was clear who had been able to return. 16 By representation we do not mean a statistically representative sample. Attendees at CCII selfselected-i.e. anyone who wanted to could come. Public participation events are typically challenged at bringing in lower income residents and renters. A key measure of success, then, is whether a sufficient cross-section of the various demographic groups is present to make the event legitimate in the eyes of the public. Deliberative polling, in contrast, aims at a statistically representative sample of the public chosen randomly (typically by land line telephone numbers) so that the resulting opinions can be considered a reliable estimate of public opinion (Luskin et al 2002). 17 In terms of age, youth were underrepresented at the New Orleans site and the middle-aged (45 to 64) were overrepresented. All districts were represented, with only District 5 (Lakeview) standing out as significantly overrepresented. The much publicized District 8 (Lower 9 th Ward) was represented in proportion to its pre-Katrina population, as was District 9 (East New Orleans). In terms of gender, our table observations indicated a predominance of women over men at a ratio of approximately 60 percent to 40 percent. The broad demographic representation of pre-Katrina population, underscored by the sign language interpreters on the front stage and the translators for Spanish and Vietnamese at special tables, was critical to the event's success.

Attrition: Who Left and Who Stayed?
While attrition over the course of a day-long event is to be expected, the demographics of attrition during the day can reveal whether certain groups left the conversation at greater rates than others. Our table observations, which included those not using their keypads, showed that about one out of three people (34%) left before the last discussion cycle in mid-afternoon. Whites had a higher attrition rate than blacks, and men had a higher attrition rate than women. The highest attrition was among white males: one out of two white males left before the last discussion cycle. Half of the white males left early Black women over 65 were most likely to stay.

Inclusion and Voice: Who Spoke?
Who spoke the most frequently? A crude but simple way of measuring participation rates is the number of times spoken by individuals in different demographic groups. This measure does not capture the length of time spoken, the content, nor the impact of the intervention; nevertheless, it gives a picture of whether there were significant imbalances in the table deliberations among different demographic groups. 18 Our table observations show that while on average there was active participation in the discussions (average 6 times spoken per person in a 15-30 minute conversation), not all demographic groups contributed equally.
Whites spoke more frequently than blacks Females spoke more frequently than males White females spoke significantly more frequently than any other cohort. Table 3 shows the average number of times spoken per person by race and gender across the 48 table deliberations observed (3 deliberative rounds at each of the 16 tables): Table 3 Source: Observations of times spoken at each of 48 conversations at 16 tables.
Who spoke the least often? Another indicator of whether the playing field was level across race and gender at the table discussions is the relative demographic composition of the participants who spoke only once or not at all in a deliberation. 19 The results show significant differences by race (Table 4): One out of four African Americans spoke only once or not at all. Only one out of eight whites spoke once or not at all. How did participation rates vary by racial mix at the table? We wanted to know whether participation rates increased for African Americans as the percentage of African Americans at the table increased. In other words, might African Americans have felt more comfortable speaking at tables that were predominantly of color? We found that the proportion of African Americans at the table did not make a significant difference in their participation rates (Table 5).
The number of times spoken per person by African Americans stayed nearly the same regardless of the percentage of African Americans at the table. Who dominated? We considered a conversation to be dominated by one speaker when the top speaker spoke more than twice as many times as any other participant. Using that indicator we found only 4 out of 48 table conversations to have been dominated by one speaker, two of whom were white females, one a black female, and one a white male.
Less than 10% of table conversations observed were dominated by one speaker.
Keypad Use: Who Voted?
Not all the participants used the keypads. The highest keypad vote count at the New Orleans site was just over 1000, yet reported attendance there was over 1500. Therefore it can be concluded that no more than two thirds of the attendees in New Orleans were using the keypads. Our table observations reflected a number of instances where participants, particularly those over 65, were confused about the instructions, how to use the keypad, or what was being voted on, or had lost interest in voting.
No more than 2 out of 3 attendees used the keypads.
What Was Decided?
The key pad results showed quickly and clearly where there was majority agreement in the room: Flood risk. A majority of keypad voters gave two of the original options on reducing flood risk a high level of support: creating and enforcing standards to minimize risk (71%) and providing financial incentives to individuals to reduce risk (64%). One option from the tables received a majority vote and a round of applause: Category 5 levees have to be built faster, regardless of what homeowners do (58%). 20 Neighborhood Stability. A majority of respondents gave a high level of support to two of the original options: providing financial incentives for homeowners to rebuild near each other (65%) and providing information to help homeowners make their own decisions about where and how to rebuild (62%). One option from the tables received a majority vote: provide incentives for homeowners to buy blighted property in their neighborhood quickly and easily (57%).
Infrastructure. One of the official options on rebuilding infrastructure received high support from a majority of voters: concentrate available funds in areas of the city with greatest need (67%). Because the participants expressed confusion after the vote was taken over the meaning of ‗greatest need' (greatest damage, greatest population return, or greatest poverty), the city-wide planning team later announced it would disregard the results from this question.
Rental and Affordable Housing. A majority of participants gave two of the original options on affordable housing a high level of support: providing financial incentives to developers (75%) and funding low and moderate income public housing (53%).

Schools and Health Services.
A large majority of participants (73%) favored the option of reopening and rebuilding public schools and health centers based on repopulation and recovery rates. A majority also strongly supported the idea of multipurpose school facilities to reduce costs. The need to improve the quality of New Orleans' schools emerged as the most important option from the tables.
Police and Fire Services. A majority highly supported the option of building facilities according to greatest need (71%). Because confusion arose over the meaning of ‗greatest need' after the vote was taken on police and fire services (as with the infrastructure question), the city-wide planning team announced later that it would disregard the results from this question as well.
From a list of the 16 most important options that had emerged throughout the day, 14 of which had been generated at the tables, participants selected the five that they felt were the most important to rebuilding the city. 21 Within a few days of the event a summary of the keypad results was posted on the UNOP website. Each attendee received a copy in the mail, as well.
But the real meaning of the event for those attending was not contained in the keypad votes alone.

The Meaning of the Event for Participants
The participants came to give input; they came because they cared about the future of the city; and they came to learn. At the end of the day CCII participants left with a sense that something far more important than just the keypad votes had transpired. According to our exit survey, participants felt that three clear messages had been given: New Orleanians are united and can work together for the good of the city; We want to come back and rebuild; and We want to be part of the decision-making--counted and heard. Three out of four respondents from the exit survey felt more optimistic about the future of New Orleans than they had before CCII.
Respondents were delighted with the chance to talk with others, express their opinions, come up with their own options, connect with people at the diaspora sites, get immediate feedback from their votes, and make a contribution to the recovery plan. Because the forum included a wide range of demographic groups, including the diaspora population, participants considered the results of the meeting to be representative of New Orleanians as a whole. Almost all affirmed the objectivity of the theme team in condensing their table reports and reporting back new options. 22 21 The only option receiving over 10% of the vote was the one about strengthening the levees faster. 22 The theme team synthesized options and comments coming in from over 250 tables across the 21 sites, distilling them into lists of 9 to 12 options, including the original three. Participants then used the key pads to choose their top three or four and very quickly saw the collective results on a big screen, Almost everyone interviewed liked the design and format of the day, especially the table discussions. They valued the chance to discuss the issues with fellow citizens in an organized way, to hear others' points of view, and find commonalities. They felt free to express themselves, grateful to be heard, and relieved that the meeting was not politicized or controlled by the most vocal or by special interests. Most participants felt satisfied with the quality of the dialogue and the choices made.
Despite a general ‗planning fatigue' in New Orleans and frustration with the slow recovery, most of the respondents said that citizen involvement in major issues facing the city should become a regular part of city governance, and that more events like CCII should be held. All 28 respondents felt that residents should be highly involved in issues and decisions facing New Orleans going forward. Respondents felt residents should attend meetings, speak out, hold leaders accountable, have more citizens on leadership committees, stay in touch with the diaspora, and continue participatory events.
The vast majority that we interviewed left feeling satisfied with the day and more optimistic about the future of New Orleans and the recovery process itself. Even those who had been skeptical left impressed and pleased. CCII was a positive experience for those attending. It successfully engaged residents in what many said they hoped will be ongoing involvement in the city's decision-making processes. The event resulted in a new level of trust in the recovery process, trust in the future of New Orleans itself, and trust in the ability of New Orleanians to come together and make collective decisions for the common good. sometimes to loud cheers from the tables where participants felt validated to see an option they had created supported by others. 23 The 28 respondents at the New Orleans site were mostly female (20), mostly African American (22), and almost all were working age (ages 26-64). They represented ten of the city's thirteen districts, and almost all had been highly or somewhat involved in neighborhood recovery planning before CCII.

What Was Left Unsaid
Our interviews, focus groups, and table observations indicate that some participants left CCII with a yearning for more: a chance to grapple with the issues in depth, work through differences, and come to a new understanding; a chance to focus on the details and make tough trade offs in terms of dollars and cents; and a chance to brainstorm and contribute to new solutions.
Participants in each of the focus groups as well as some of the exit interviews expressed disappointment that there had not been time to thoroughly grapple with the big issues and come to consensus. Our table observations showed that the limited discussion time was often enough to reveal differences but seldom enough to deal with them. The keypad vote on funding public housing, which took place after the table dialogues, largely divided along income and homeownership lines between those highly in favor and those highly against. Two thirds of the exit interview respondents said that they did not shift their views on any topic as a result of the table conversations. One of the focus groups demonstrated the volatility that remained in the room around flashpoint issues-e.g. public housing, who should come back, and the economy vs. social needs. -The vocal participants … disagreed, they listened to one another, and they attempted to better understand the issues. They expressed frustration, not with one another, but with the enormity of what they were trying to do in a very limited time.‖ (table observer).
These concerns raise the question of how much and what kinds of deliberation are needed to make wise decisions. Some New Orleanians wanted deliberative formats that would allow greater exploration of key issues, more information and detail, the means to work through the divisive issues, and the opportunity to contribute to solutions. 24 The Outcomes

Social Trust and Social Healing
The meaning of the event is described below in participants' own words from the exit interviews, table observations, focus groups, and key informant interviews. The responses exemplify how the meeting fulfilled many social needs critical to building social trust and fostering social healing. 25 The meeting served as a forum for rebuilding community and social connections: -It's wonderful to see people who want to come back.‖ -I'm seeing people I knew and hadn't seen in months.‖ -This is healing that's taking place right now, in this room.‖ -This meeting was able to jump the barrier of neighborhood boundaries and bridge them.‖ The meeting was an opportunity to create new relationships.
-The close proximity at the tables made people bond. We enjoyed each other so much the time went by really fast‖ -There were honest, authentic feelings expressed.‖ -Everyone really bonded. They didn't leave early because they didn't want to let their table down. We became loyal to each other… --The good thing was to have everyone putting in their input. To see that energy. Closest I've felt to being in New Orleans since Katrina.
-Everybody was skeptical at the beginning. No one knew each other before and at the end everyone was friends and hugging.‖ The meeting connected dispersed New Orleanians.
-People saw the other sites. We all know they are out there, but we felt them. We've known they were there in our heads, but this time we felt them in our hearts…. Really powerful.‖ -I found communicating with the cities delightful.‖ -It was wonderful to hear them talk about wanting to come home.‖ -On Saturday, we raised the bar. We cannot make any more community decisions without including the displaced citizens. That is what I am most happy about.‖ The meeting showed the ability and importance of New Orleanians to work together, no matter who or where they are.
-We all need to be unified and on the same page.‖ -We can work together for a common good;‖ -I see this as a first step -we're getting together;‖ -Important to be together;‖ -We can work together for New Orleans.‖ The meeting allowed New Orleanians to hear each other and be heard by their government.
-It was a blessing to be heard.‖ -Great opportunity to be heard individually and collectively.‖ not refer to the individual healing from the immediate trauma of disaster and displacement, which is most often addressed as part of relief and response efforts, but rather a longer term re-establishment of a collective identity.
-I like that they are trying to get input and be inclusive of as many people as possible. I appreciate the opportunity to give feedback and did so.‖ -Who is listening? Nagin, councilmen, every participant, the people who were not affected hear us. It takes the mystery out of it because they saw us and heard us.‖ The meeting was an opportunity for mutual, collaborative learning that increased the participants respect for each other and mutual understanding of the issues.
-We had a diverse group of people, but we were able to come to an accord.‖ -It was very positive and good to know there is a common understanding.‖ -People were impressed by their neighbors and strangers.‖ -Most important was to see how despite some differences… we were amazingly the same. We all have a real sense of the priorities after 14 months and where we focus our attention…. It was validating.‖ -I came to bring the neighborhood's voice to the process and be part of it myself….I was shocked by how much I enjoyed it, and by the quality of what other people had to say.‖ -The best thing that emerged for people was the chance to sit down together and talk about various issues with other New Orleanians.
Everyone liked talking about each issue and coming up with a consensus, including when they brought something new to the table that wasn't one of the choices. The spirit of collaboration was strong.‖ The meeting helped reinforce a sense of collective commitment to the future of New Orleans: -We are a community, a group of survivors, and we will stand steadfast.‖ -We believe we can rebuild. We can do it with or without a lot of help. Without help it just will take longer, but we'll do it.‖ -We are formidable and willing to go out on a limb with our own resources, our own wit, and our own backbones to accomplish what we believe we need to.‖ The meeting resulted in citizens wanting more community engagement, especially ongoing participation in the recovery process.
-I hope they will keep trying. It's important for me to hear other people respond and know what other neighborhoods were thinking.‖ -This needs to continue. These organized methods (can be put) to use in all matters. It (CCII) illuminated the need for dialogue. People have solutions but no one is listening to them.‖ -Certainly UNOP with this meeting is more effective than anything we had prior to this. Everyone is talking about ‗unprecedented.' We haven't had this level of civic involvement before. I hope … we can do a lot more of this.‖ CCII helped build social trust across racial, economic, and social divides (-we can work together!‖). It helped create a sense of future life for the community (-New Orleans will come back!‖). It fostered a sense of social efficacy (-we can rebuild‖), a sense of individual responsibility and desire to participate (-I count‖), and a sense of stewardship of the whole (-I care‖). This is social healing: CCII helped participants to re-member themselves as a whole, as the social field that is New Orleans. 26 Influence: What Difference Did the Public's Input Make?
CCII was seen by the public as legitimate because of the high attendance, the broad representation of New Orleans' diverse population, the transparency of the process, and the fact that the planners and many public officials were listening. Certainly CCII galvanized broad public support for the UNOP process (Williamson 2007). However, two doubts were raised frequently: Did the framing and clarity of the questions and options allow for meaningful input? Would the public's input be reflected in the final plan and in actual implementation? Participants wanted to know if CCII was just an exercise or would their voices count.

Did the framing and clarity of the questions allow for meaningful input?
The concern about the vagueness, -dumbing down,‖ ambiguity, and misunderstanding of questions was a theme across all of our sources. Some respondents found the questions and options confusing and did not know if their vote matched the intention of the option. They were particularly frustrated by the ambiguity of -greatest need‖ in two of the options. Some found that there were so many priorities discussed that the directives to the planners would not be clear. 27 Others found the questions too general (-shotgun‖) to be useful. The process of framing the questions was questioned in our key informant interviews. Some felt that different stakeholder groups had not been adequately involved in the development of the options.

Would the public's input be reflected in the final UNOP plan and implementation?
The city-wide planning team, which had left the participatory process to Concordia and AmericaSpeaks as much as possible in order to focus on the technical studies, got the message loud and clear that the final plan would have to readily demonstrate the public's input if it was to go forward. While not being advocates of deliberative planning, 28 the team went to work to incorporate the messages of CCII in time for the plan's presentation to the public for approval just six weeks later at Community Congress III. The team gained some immediate goodwill by announcing that the two questions about -greatest need‖ would be disregarded because of the ambiguity in the wording. 26 The term ‗field' is borrowed here from dynamic systems theory to describe the connection among the members of a living social system (Scharmer 2007). 27 Three out of four exit interviewees could identify only one clear vote: build better levees faster. 28 Source: personal communications with three of the principals prior to CCII.
The UNOP plan presented at CCIII reflected back to New Orleanians not just a literal interpretation of the keypad results but an understanding of the sentiments underlying them: safety from future flooding; empowering the community to act; an opportunity for all to return; equitable public services; and quality schools that meet community needs. The plan focused on empowering homeowners through incentives for clustering homes on higher ground. It proposed voluntary standards for rebuilding, including financial assistance for elevating homes beyond FEMA requirements and information for flood risk mitigation. It recognized the desire to make New Orleans affordable for low-income families, public housing residents, and renters, while at the same time avoiding concentrations of poverty and violence by building mixed-income communities. The plan called for targeted infrastructure investment, including neighborhood-based health centers and clinics, and multi-purpose school facilities. It identified implementation strategies based on flood risk and repopulation rates, and set time frames. It called for fairness and equity by proposing a funding gap program to meet the true costs of elevating or rebuilding. The public could rest assured that their concerns had been reflected in the UNOP policy plan. follow through with continued opportunities for public deliberation and oversight of the recovery process, that the ‗big shot planners' would go it alone after the event, that the mayor would ignore the plan, or that special interests or sheer incompetence would derail the plan. 31 If there was no visible follow through by the Office of Recovery Management, the halo of CCII and UNOP would quickly erode.
By September 2007 the promised ‗cranes in the sky' had not materialized. Public faith in the recovery process began to erode. In December one of the flashpoint issues left unresolved at CCII reared its head: the fate of the public housing projects. With no deliberative mechanisms in place to address the issue, Federal bulldozing of public housing sparked loud protests that turned violent at a city council hearing where the vote to bulldoze was known ahead of time. Where was the new New Orleans? By the end of 2007, the necessary institutional infrastructure for ongoing civic engagement in the recovery process had not been built.
To be self-sustaining, participatory mechanisms must be integrated into the City's institutional infrastructure at the neighborhood, district, and city-wide levels. 32 The participatory mechanisms must include not only the one-way information flows of public information campaigns and public hearings, but also interactive mechanisms for deliberation and dialogue. Funding for the participatory infrastructure cannot depend in the long run on ad hoc sources. Private foundations have funded the deliberative efforts in New Orleans thus far. In fact, according to a GNOF representative, private foundations salvaged the recovery planning process in New Orleans by courting the key stakeholders to get them on the same page and financing the participatory process necessary to make the plan legitimate. The dividends were great-but temporary. A pivotal point on the road to deliberative democracy for the new New Orleans will be passing the impetus for public deliberation from private foundations to the public sector. 33

Summary and Conclusions
CCII was a remarkable event in a challenging setting on the road to participatory democracy. In a city plagued by racial divisions, economic disparity, and the trauma of natural disaster, the event demonstrates that inclusive public deliberation does more than provide reasoned input and a public voice into difficult policy decisions. It 31 Polletta's results from AmericaSpeaks' -Listening to the City‖ event in New York show similar fears among participants that the public's voice would go unheeded (Polletta 2008). 32 It is important to build the institutional framework of participation around the district and neighborhood planning capability put in place by UNOP and the various neighborhood initiatives. The framework should incorporate recovery monitoring mechanisms at the neighborhood and district levels. It should also facilitate small business and homeowner reconstruction, especially in targeted areas, through localized assistance centers for planning, designing, financing, and building. 33 Foundation support may be critical in kick-starting public deliberation and funding prototypes; but if the events stay at the level of ad hoc or pilot projects, the public may soon experience ‗participation fatigue,' the feeling that ‗this is just an exercise.' As a result people may turn to apathy, self-interest, power plays, or protest. does more than legitimate new public initiatives. It can spark a sense of common purpose, connect one another through a shared love of place, and rekindle faith in the future of the beloved community.
CCII was the pivotal point in galvanizing public support for the New Orleans recovery plan. It gave voice to the public's priorities, which were reflected in the final UNOP plan. It helped New Orleans re-member itself, across racial, economic, and geographic divides. It will be up to the City to consolidate these gains through a transparent institutional infrastructure for ongoing participation linked to decisionmaking and action.

Lessons for Deliberative Democracy
Elements of good process design. Effective deliberative meetings do not just happen. They are designed, planned, organized, facilitated, and followed-up, and they are embedded seamlessly in a larger process of public decision-making and participation. CCII illustrates several key elements to good process design: AmericaSpeaks mobilized an unprecedented outreach campaign to assure strong representation of hard-to-reach demographic groups, especially the low income, renters, and diaspora populations. The meeting design incorporated procedural fairness, which gave a sense of safety and freedom from political manipulation. 34 The event design appealed to both head and heart: it incorporated art, poetry, music, humor and drama, as well as information and deliberation. Facilitation of the table conversations helped prevent domination of the conversations by one or two speakers. Conversation was not limited to the official policy options: new options and minority opinions could be expressed and reported. The theme team process for synthesizing table input and the rapid report-back of keypad results created transparency. Followup after the meeting was immediate, with results posted on the website and mailed to individual participants.
The UNOP process, in which the event was embedded, also exhibited elements of good process design. It maintained transparency and credibility through clear communication with the public, quick and easy accessibility of information, and follow-through on stated intentions. It offered multiple opportunities for public participation at different levels (neighborhood, district, and city-wide) and different tracks (community advisory committee, district steering committees, stakeholder groups, and general public). The results of CCII were clearly and rapidly incorporated into the UNOP plan and the public given the opportunity just six weeks later at CCIII to see the plan and respond. The plan garnered a strong endorsement by over 90 percent of those responding on the keypads.
Ongoing challenges in process design. Table observations at CCII revealed two ongoing design challenges for deliberation professionals: first, how to make the table conversations equally appealing and comfortable across race, culture, and gender. At CCII women spoke more frequently than men and whites spoke more frequently than blacks. White females had by far the highest participation rate per capita. White males were most likely to leave the event early. Making the use of technology equally accessible across age and education is another design challenge: at least a third of attendees did not use their keypads for voting.
Feedback from CCII also reveals another design challenge: one size does not fit all. Different issues, needs, and communication styles require different process formats and designs. A good public participation strategy will incorporate an array of meeting types. In the case of CCII some participants were left with the desire for a deliberative format that would allow greater exploration of key issues, more information and detail, the means to work through the divisive issues, and the opportunity to contribute to solutions.
Recognizing the influence of deliberation in social trust and social healing. The very act of bringing citizens together to make decisions is an opportunity to build social trust and social capital. CCII was a catalyst for connection across the tables and across the miles. It helped build social trust across racial, economic, and social divides (-we can work together!‖) and it fostered a sense of social efficacy (-we can rebuild‖). But something more happened at CCII: New Orleanians saw the social field that is New Orleans reconnected for the first time since Katrina. CCII helped create a sense of future life for the community (-New Orleans will come back!‖). It cultivated a sense of individual civic responsibility and desire to participate (-I count‖), and a sense of mutual respect and stewardship of the whole (-I care‖). It helped residents re-member New Orleans in their hearts and minds. This was social healing: a moment of overcoming isolation and becoming whole. Not just feel-good byproducts of public deliberation, social trust and social healing are important dividends of deliberative democracy, especially in communities on the road to recovery from natural disaster.
The need for an institutional infrastructure for public participation. Deliberative public events, like most public events, have a short half life. Their momentum can be sustained only by embedding them in an institutional infrastructure of civic engagement. Such an infrastructure creates consistent opportunities for public deliberation connected to decision-making. It builds a deliberative feedback loop between government and citizens. Done well the infrastructure of participation is designed with the stakeholders and institutionalized over time to include the necessary tracks, levels, formats and elements for different communication styles and purposes. It builds an outreach capability that reaches a broad cross-section of the public. It involves stakeholder representatives and the general public in multiple tracks. It engages the public at the city-wide level and the neighborhood level. It encourages and incorporates neighborhood initiatives, linking participation to action. A civic engagement strategy for long term disaster recovery is as necessary as an investment strategy. With a strategic framework, an institutional infrastructure for civic engagement can evolve gradually through city-launched prototypes. Deliberative democracy is not a series of ad hoc events. It is a way of governance.