Our national, state, and local government systems serve more people that were imagined at the point of design. North Carolina, for example, has more than twice as many people as the entire United States in 1790, the year before the state ratified the U.S. Constitution.
On good days, I admire the governmental chassis the early engineers designed and remark at how well it has scaled. On my more pessimistic days, I worry that our structures for governance, politics and public policy are buckling; that our trust in institutions is declining; and that our democratic feedback loops are cracking.
Take one of the bastions of our democratic feedback loop: the public hearing. Whether legislative or executive in nature, public hearings are often political charades designed to facilitate the image of democracy. Policymakers with minds made up and political agreements already reached sit silently during public presentations only to pivot quickly down a preordained path without any consideration of the public comment.
It is hard to blame the policymakers for this posture. When they do listen and try to fully engage with their constituents, the outcome is often cacophony. The size of many constituencies makes the volume of feedback onerous. Furthermore, the people who are often the most engaged are not representative of a majority; they are simply the most affected.
The good news is that we can now harness social media technology to reinforce and even improve the feedback loop.
A New Kind of Feedback Loop
Last week, I gave an overview of some of the social media, messaging and chatbot evolutions. What if these evolutions could be harnessed to engage citizens at all levels? What if the engagement was based on a two-way conversation? And what if it resulted in more informed decision-making for a community? Here is a hypothetical example of how the technology could work:
A town is considering whether to change its trash collection process. The town staff is considering different options that would enhance recycling and slightly raise the cost of waste removal. There are highly-motivated minorities of citizens in support of and opposed to the plan who will turn out in force to voice their opinion, but 95 percent of the population does not have a strong opinion.
Interested in gauging broader public opinion, the town launches a chatbot designed to engage the public about their waste habits. The experience involves a five-question, multiple choice survey and includes a couple of open-ended questions that allow for people share their thoughts in free form. Links to the chatbot survey are distributed through multiple media: Facebook, email newsletters, print advertising, and billboards that simply say, “Thoughts on Trash? Text 555555# to tell us what you think.”
Assume 100 people respond in the first day and 10 of the respondents mention composting in their open-ended response. The town can update the engagement tool to add a composting question so future participants can help provide more compost-specific data.
Further assume that one of the respondents provides critical feedback about the existing service and suggests a problem with their current trash service. The town then can engage via text with that person and ask for more specific information about their problem. The participant can choose whether or not they want to engage and provide additional information. If they do not, the interaction ends. If they do, they can get customer service beyond what they might have otherwise requested. It is a win-win for the town.
Real World Examples
The Center’s sister organization, EducationNC, has been test-driving this technology in exciting ways. The platform Reach NC Voices, is already enhancing EdNC reporting. Take its recent coverage around end-of-grade tests and final assessments. To supplement the reporting, the Reach NC Voices team engaged with readers about their experiences. EdNC reporter Liz Bell, Managing Editor Laura Lee, and the Reach NC Voices team distributed a set of questions about testing online and via texting to teachers and parents across North Carolina. The initial responses shaped the first story but will also inform subsequent reporting.
Reach NC Voices, as a platform and initiative, offers a unique opportunity to power EdNC and NCCPPR journalism and research and serve as the most robust civic engagement platform in North Carolina.
In Detroit, a group called Outlier Media is using a similar approach. Focusing on consumer protection and predatory financial practices, the self-described journalism outlet does not publish articles and only communicates directly with readers via text. It is one-third investigative news, one-third whistleblowing sight and one-third data analysis. Using their social media tools, Outlier uncovered a story about a dubious rent-to-own program. Outlier reporters used the technology to both gather and share data about the problem with affected residents. The story later appeared in the Detroit News.
With the Reach NC Voices, EdNC is testing ways to use the technology. Here are some potential platforms that the team is exploring:
Microtargeting involves engaging with a small population of people with a clearly defined association—small town psychiatrists, for example—to understand their perspective. The engagement begins with a short survey/poll but evolves into open-ended questions and to a two-way conversation. If the information is newsworthy, it provides an initial foundation for publication. Regardless, it provides useful information for policymakers, employers or others that focus on the group or issue.
This kind of microtargeting combines the power of traditional surveys but has a much lower barrier to entry. The social media technology allows for the connection to groups of people that may be difficult to track down or assemble, as well as groups that are already assembled. Furthermore, the costs of implementing the technology are relatively low.
Many news stories and policy initiatives begin after reporters and policymakers have had a chance to listen to what is going on. Some of the best of these interactions are informal and initiated without a clear goal in mind. The broad goals are to listen, to ask questions, and to follow-up in order to uncover connected pieces of information.
The social media technology allows for a kind of virtual porch sitting, as a friend of mine calls it. Unlike actual porch sitting, these engagements can happen both in real time and asynchronously. In addition, the technology allows for people who might not usually be engaged to be part of the conversation, which opens up new porches, trends and potential issues to be explored down the line.
Politicians like polls and focus groups. They are some of the best ways to get real information about the way people think. At this time, social media driven engagements will not replace polls, which are designed by social scientists and statisticians to be representative. However, these sorts of engagement tools can be used to gather non-scientific feedback. This feedback may not rise to the level of solely supporting a news story or policy initiative, but it can give policymakers and others a sense of the mood on an issue and explore what questions still need to be asked.
Reach NC Voices lead Nation Hahn described this sort of engagement as demonstrating “the power of open-ended questions.” It allows for hunches to develop into news leads and policy ideas to evolve into solutions. He also clarified that the technology is evolving to more closely replicate sampling. Reach NC Voices and its partners are developing demography weighting tools that may come closer to the accuracy levels of polls.
In good times, governments, employers and other administrative bodies struggle with how to communicate effectively with their constituents. Following a hurricane, flood or other crisis, engagement can be even more difficult. After Hurricane Matthew, the state and local governments faced a dispersed population suffering from a wide range of needs. In addition, they were struggling with providing information and coordinating assistance, as well as combatting misinformation that spread quickly.
Many employers and administrators have been implementing one-way communication systems for several years. The University of North Carolina and NC State University, for example, regularly text employs about safety concerns, weather alerts and similar events. That sort of engagement is a great initial step, but technology systems like Reach NC Voices can help communicate and gather information in ways that can enhance policy and other administrative decision-making.
Amidst the great potential of this technology, it is important to note a few cautions.
- The feedback loop is as only as good as the analytical engine that gathers, measures and responds to the data. The technology platforms allow for relatively easy data collection and analysis of close-ended questions (i.e., multiple choice). Open-ended questions require more sophisticated systems, both technologically and in terms of human analysis.
- There are saturation points that must be balanced. As more entities utilize the technology, the engaged public will begin to be inundated with people looking to engage. The Outlier platform in Detroit is novel now, but will it be in three years where there are 10 different issues around which people can text and engage?
- It is tempting to use these engagements in the same way that we use polls and other social science. As we have seen in recent elections, even the social science from formal polls can be unreliable, and the best of those are designed using sophisticated statistical methodology. For the most part, there is little to know scientific process that goes into these polls. Furthermore, they lack the random design that is meant to simulate a true representative sample.
- Because these engagements come via chatbots and other computer programs that simulate a two-way conversation with another being, there are potential dangers in overly personifying technology. Much has been written about the dangers of dehumanizing individuals and how that can be used to perpetuate societal bad acts. I wonder if the contrapositive is also true: humanizing software programs has the potential for individuals on both sides of the engagement to shirk some of the responsibility for their interactions. Make no mistake: there are people behind all of these bot technologies. Some are trying to steer your interests and consumer habits. Some are trying to avoid having to spend the money to have a human interact with you. Others are hoping to avoid your questions and frustrations all together. In the mix of all of this, there is an opportunity for genuine engagement and to enhance how we interact with each other and our communities, but we need to be mindful of how the software interfaces affect our social mores.
Technology and journalism expert Clay Shirky wrote, “Communications tools don’t get socially interesting until they get technologically boring.” We are easing into the period where the social media technology seems to have stabilized and is now getting interesting. Technologies that were once considered the purview of millennials are now broadly used across the population. According to a Nieman Labs report, 85 percent of the population gets some news from mobile devices, including more than two-thirds of people older than 65. This is no longer a young person’s game.
The possibilities have high limits. The guardrails for engagement are still being set. Now is the time for journalism and public policy to accelerate the exploration of these technologies so that they enhance our community and not just tell us which Walking Dead characters we resemble.
My friends at EdNC are doing their part. Stay tuned. And while you are at it, go ahead and engage in this new public square.
This questionnaire covers a broad set of topics to help us more completely understand beliefs about policy and governance.
Andrew Holton is a board member and contributor to the N.C. Center for Public Policy Research.