Capturing the Wisdom of Crowds

Combining citizen intelligence and online civic platforms.

By Kevin C. Desouza and Kendra L. Smith

Online civic platforms are all the rage today. Examples include: CoURBANIZE, MindMixer, CivicLab, Neighbor.ly, and Civinomics. They are emerging at an ever-increasing rate, all with the simple goal of harnessing collective intelligence to solve complex social problems and discover opportunities for innovation. Robert Steele, the author, activist, and advocate for open source intelligence, noted that citizen intelligence makes "possible the revival of the collective will of the people."

The Internet, he said, "is an instrument of frontline communication, largely uncensored, for those individuals who wish to communicate, calculate, and inform themselves and each other and, in so doing, perform their responsibilities as citizens of the Republic."

More than one observer has said that the greatness of crowds lies within the richness of knowledge available among many members of the group.

Francis Galton, in 1907, was one of the earliest writers to address the wisdom of crowds. He conducted a study that asked 787 participants to guess the weight of an ox. One group of participants were experts, such as butchers and farmers, and the other group were men with no special knowledge of the matter. The group of nonexpert guessers had an average guess of 1,197 pounds; the actual weight of the ox was 1,198.

The civic platforms that leverage collective intelligence to solve policy and planning issues come in varied shapes and sizes. Over the last few years, we and our colleagues have studied the features of online civic platforms in major metropolitan areas within the U.S. and Europe. One feature is common across the range of platforms: Engage the masses through technology to solve problems and realize opportunities that have an impact on all those who participate on the platform.

At the Massachusetts Institute of Technology, the Climate CoLab is a collective intelligence platform developed to address the problem of global climate change. It is an online platform of nearly 4,000 people who generate proposals. Users (anyone can join the community and participate) are invited to submit proposals and comment on others. Users collaborate regularly with those they have met on the site to develop proposals.

Motivations for citizen participation

Citizens are getting involved because technology has enabled us to do more while going about our normal lives. Someone with a smartphone or tablet can easily contribute to projects. Outside of San Francisco, at Mount Diablo State Park, hikers can take a photo at a marked hill and upload the photo to Flickr, Twitter, or Instagram using the hashtag #morganfire01.

The pictures go into a bank that monitors that area's ecological recovery after a wildfire. Researchers compile the photos into a crowdsourced time-lapse series that shows the process of vegetation growing back. The information from the project helps researchers determine the long-term impacts of wildfires.

Clay Shirky, a researcher and consultant on the social impacts of the Internet, suggests that cognitive surplus allows us to collaborate and make the world better. He says cognitive surplus is made up of two things: individuals' free time and talents and the mediated landscape of today. You might be thinking that people with talent and free time have existed forever. However, free time and talent, coupled with today's technology, make individuals more likely to engage because it is relatively easy to connect.

An example: Following the 2007 presidential election in Kenya and the crisis that erupted thereafter, Ory Okolloh, a Kenyan activist and lawyer, began blogging about events. She started soliciting information from others — and it began pouring in. She noted on her site that the amount of input was becoming unmanageable and she wished there were a way to automate things. Two programmers read her blog and did just that. They set up an open-source technology called Ushahidi that allows for anyone to gather information to find solutions. The platform is now widely used in several countries, including Haiti and Chile, where Ushahidi was used after the 2010 earthquakes to gather information and assist in humanitarian efforts.

Citizens are also motivated to share and transfer information as a matter of civic duty. This impulse may not be particularly strong in the U.S., but some problems cannot be ignored. Citizens are dealing with the effects of terrorism, recession, poverty, crime, natural disasters, and unemployment, and they feel compelled to improve their quality of life.

After the 2013 Boston Marathon bombing, law enforcement asked citizens to upload photos and videos that might aid in the identification and capture of suspects. In other pockets of the Internet, citizens were already doing their own investigations by uploading information to online communities such as Reddit.

In Waynesboro, Virginia, police were unable to identify the car involved in a fatal hit-and-run. The police department released a photo and description of a metal piece left at the site to see if the public could help identify it. Soon thereafter, a popular automotive blog picked up the information and commenters provided the information needed.

The blogger contacted the police department with the information, which led to the positive identification of the vehicle and the conviction of the perpetrator.

Archetypes of civic platforms

Technology platforms for citizen intelligence are springing up quickly. Platforms such as Deliberatorium, DebateGraph, Cohere, YourView, and CoPe_it! all allow for extensive discourse. Each has special features such as multiple ways to contact other users and participate in discussion boards. Additionally, these platforms employ social analytics, discourse analytics, and social network maps. These sites allow users to gather information and debate ideas and solutions to specific community issues.

Users can also add evidence and information to other users' claims, which triggers conversations and sharing. In many U.S. cities, leaders are finding value in citizen intelligence. Online civic platforms tend to fall into four main categories, as one of us has also noted in an upcoming Journal of Urban Technology article.

Model 1: Citizen-Centric and Citizen-Sourced Data

Citizens are the principal actors on these platforms; they create and set up the platforms themselves, engage their community members, build the rules of the platform, and moderate content. To engage with citizens, they can create group discussions on thematic issues, set up competitions to crowdsource ideas, and request votes for selection of ideas — or they can just be a repository of key concerns and information of interest to the community. They try to help solve local governance challenges.

Local government has no formal role is this model and can only exert inflence passively, for example by receiving information and solutions once they have been chosen and vetted by the citizen users. Governments are not bound to implement the solutions.

Change by Us Philly is an online platform for community projects that was launched by CEOs for Cities, Local Projects, and Code for America in collaboration with the city of Philadelphia. On this platform, citizens can suggest changes they would like to see, join grassroots projects, or create small groups to discuss issues. When citizens submit ideas, the ideas are reviewed and, if approved, are passed along to city leaders.

Similarly, Localocracy services the Massachusetts cities of Arlington, Cambridge, Granby, Milford, and South Hadley by allowing users to discuss local issues and generate and select ideas. To engage, citizens must register with their actual identities (to ensure that they are registered voters); then they are allowed to view their leaders' and neighbors' rationale for or against a particular concept.

Once an idea is selected, they propose the solution to the appropriate public agency. The local governments also play a passive role, but Localocracy invites public agency representatives to actively monitor the suggestions. Implementation is at the discretion of the agency.

Further afield, citizen groups in Kampala, Uganda, are testing an anticorruption app currently in development. The app, called Action for Transparency (A4T), allows citizens to see how much money schools and health centers are allocated, for what reasons, and how much money is being spent, or not being spent, properly. When citizens see signs of impropriety, they can blow the whistle and notify their partner, Transparency International.

Model 2: Citizen-Centric and Government's Open Data

As with Model 1, citizens are the principal actors. The focus of these platforms is to harness open public data. Unlike Model 1, open data here is the main attractor and the interaction happens between citizens and the open data. There is limited citizen-to-citizen or citizen-to-agency interaction. Citizens typically use simple tools to download information and explanations about the datasets as well as discussion groups and forums where they can post questions and seek thematic communities that are working on similar datasets (for example, crime or housing data).

On www.data.gov/cities, any user can retrieve the city of Seattle's 911 fire calls in real time. All the user need do is filter the information (by day, by medical or fire incident, by location) and download the data in a number of different formats (CSV, PDF, RDF, JSON, RSS, XLS). Users can also post comments and dialogue with other open data users. Finally, users can view information about the data: how many times it has been downloaded, licensing and attribution, and information about the data owner.

A citizen developed the NYC Health Ratings website using restaurant inspection data from the New York Department of Health and Mental Hygiene. The data offers citizens easy access to health safety ratings of restaurants. In Oakland, California, the Oakland Crimespotting website was developed to provide citizens with up-to-date information about criminal incidents. Using open data on crime incidents from law enforcement agencies, the website displays information on an interactive map. This website allows citizens to track crimes based on type and to leave comments.

In Chicago, the Transit Future app is a campaign to encourage the city to build new rail lines. It is an interactive map that explains the future of transportation in the city using city data.

Model 3: Government-Centric and Citizen-Sourced Data

In this model, public agencies take the lead in developing citizen intelligence through participatory platforms. Unlike Models 1 and 2, this model involves substantial interaction between local governments and citizens.

This platform generally allows the public agency to set and control the narrative. Citizens can hear and be heard, but they lack the proper channels to ensure that their opinions and feedback are considered. In other cases, the public agency acts as a near equal and the information flows in both directions. Citizens can communicate with public agencies quickly. However, idea implementation is still at the discretion of the agency.

Austin, Texas, developed SpeakUpAustin to engage citizens on local issues. City agencies identify problems, make those problems available for citizens to read, and ask citizens to respond. They also can vote on proposed solutions. After a solution has been voted on, the application idea portal remains open to keep citizens updated on progress. The city of Wichita used MindMixer during a drought to ask citizens what they would be willing to do to conserve water. The city and county of San Francisco used ImproveSF to call for library card redesigns and received more than 5,000 submissions from the community. The site also tackles several issues such as transportation, healthy living, and community development.

The Regional Transportation Commission in Washoe County, Nevada, developed an online web map and app for the Reno Sparks Bicycle and Pedestrian Master Plan. That plan seeks to create a comprehensive system of bicycle and pedestrian routes. The program solicits resident opinions on new locations for bike trails. Along with public meetings, citizens can take photos of a bike lane, curb, or intersection and submit a comment to the web map such as "needs crosswalk" or "nice scenery." The comments are displayed on an online map.

Model 4: Government-Centric and Citizen-Developed Solution

In this model, the public agency develops a website platform for citizens to devise actual solutions such as mobile apps. As with Model 3, the public agency is still the primary actor; agencies seek and encourage citizens to become citizen software developers, using tools such as open data and mentorship to create a solution.

This model offers a true partnership because it is government centric but with a high level of citizen power. It requires a lot of collaboration since the public agency expects to use the citizen outputs as a mainstream technology. When complete, information is exchanged with the public agency in the form of a website or app.

Many times, the agency will offer monetary prizes to encourage participation. Other lures may include coaching, mentorship, and training. New York City has hosted the New York Big Apps competition for five years. The competition set out to have groups use city datasets and create novel applications with the information. A panel of judges voted on the entries, and the city offered developers a few different incentives: money, three months of workspace, meeting the mayor, and access to organizational connections.

This year, the United Kingdom was battered by storms that led to mass flooding. In response, the nation's Environment Agency called on developers to create solutions with government data (including flood sensors) in a competition called Flood Hack. Here are a few solutions that resulted: Flood Feeder aggregates flood information such as warnings, alerts, mobile phone mast locations, and transport routes. Citizen Flood Journalism and Fludbud help to connect people tweeting from flood-affected areas. FloodRelief connects volunteers, local authorities, and emergency services.

Moving forward

Encouraging citizens to use their talents and skills to help grow and improve their city and themselves is the promise that lies ahead in using citizen intelligence. Some suggested guidelines for planners and public officials:

Learn more about citizen intelligence platforms. Each has different capabilities, and there is sure to be a platform that's right for your organization.
Investigate each platform to understand which work well and how to partner with them. Eventually, this will become a very crowded space and only the most relevant and engaging platforms will survive. Technology is in vogue one day (BlackBerry, MySpace) and out of the mainstream the next. Finding the right platform is paramount to success.
Try to build links and connections to those who run these platforms. The operators truly believe in the power of citizen intelligence and want to engage with as many entities as possible.
Figure out which problems can be outsourced to these platforms. This means understanding the platforms' capabilities. Traditionally, many planners or researchers avoid engaging citizens in difficult tasks or decisions, but in this space it is warranted. High-level platforms are capable of facilitating detailed, in-depth conversations.
Understand that you are seeking intelligence on community issues. You can end up with unwanted or undesirable outcomes. You will not be able to control the narrative, so your city must accept early on that citizen ideas might not align with city constraints or priorities. Remember, citizen intelligence focuses solely on the public's interest. Local government must also be concerned with job security, political agendas, and budgetary constraints. If you use these platforms, you must be willing to go the distance no matter the outcome.

Social problems are messier than technical problems and not easily fixed. Consider homelessness. This is one of world's largest problems — a huge issue that concerns advocates for human rights, economic development, housing, community development, social services, law enforcement, nonprofits, and health care. This issue cannot be solved by public meetings alone.

In hopes of reaching a quality outcome, Christian Sarkar, a strategy consultant, developed the concept of the $300 house to help improve housing for the world's poorest and help break the cycle of poverty. In a 2011 open design challenge, experts in design, energy, finance, and urban planning worked to develop and sharpen ideas on how to execute a $300 house.

Still in the early testing and experimentation stages of the project, the developers are finding how expansive their work truly is. This is not just about houses; issues such as pollution reduction, extreme unemployment, health, sanitation, sustainability, government relations, and policy are all involved. Collective intelligence platforms offer "netizens" the opportunity to engage in important civic discourse as well as develop solutions collaboratively.

Kevin C. Desouza is the associate dean for research in the College of Public Programs, a professor in the School of Public Affairs, Senior Research Fellow at the Center for Organization Research and Design, and the interim director for the Decision Theater in the Office of Knowledge Enterprise Development at Arizona State University. He is also a nonresident senior fellow at the Brookings Institution. Kendra L. Smith is a doctoral candidate in Arizona State's School of Community Resources and Development.