Brownsville Sentiment: Equity Project

Public Sentiment, in a cross-sector collaboration with the City of New York and NYU-CUSP, is using data and civic engagement to help residents and decisionmakers develop effective strategies for immediate pandemic relief, and long-term recovery. 

Our approach involves data analysis on existing City-wide data, a year-long, bi-lingual, bi-weekly survey to track residents’ evolving needs, resident-led data analyses, and iterative, collaborative policy design with policymakers. 

This work, adaptable to any global context, is:

1) Immediately connecting residents to appropriate relief organizations and government agencies, based on their specific challenges (eg: disruption of social and public services);

2) Informing policy makers and researchers about the mental, physical, and behavioral effects of a pandemic to help inform future prevention, relief, and recovery strategies

3) Sharing residents’ insights to help decisionmakers in local government and organizations design and deploy relief programs and policies.

We’re working to slow and prevent the spread of global diseases like Covid-19 in lower-income communities in the U.S., and around the world.

In communities everywhere, the Covid-19 pandemic has created new and urgent challenges, and exacerbated deeply-entrenched challenges long in need of systemic reform. It has highlighted disparities in access to quality healthcare, reliable information, and use of effective preventative practices. These disparities fall largely along racial and socio-economic lines. 

Within New York City, the global epicenter of the pandemic, communities of color in the boroughs of Brooklyn and Queens have been disproportionately hard. In Brownsville, Brooklyn, where our program runs, hospitalization rates for COVID-19 are twice that of the City’s average. 

There are approximately 112,000 Brownsville residents. 71% of residents identify as black, 20% as Hispanic. Before the pandemic, the community was dealing with a 16% unemployment rate. Nearly 40% of residents live below the poverty line, with median household income of $21,000. This is nearly 67% less than the citywide median household income.

The trend of lower-income communities being hit harder than their wealthier neighbors is not unique to New York City. This is equally true in states across the U.S. and in countries around the world. 

Our approach to Covid-19 recovery involves:

• Online survey co-designed by neuroscientists, social psychologists, and Resident Working Group
• Survey runs for at least one year, is bilingual (Spanish and English), released every 2 weeks to track evolution of needs and outlooks; coordinated distribution through CBO network, government agencies, including NYCHA
• Resident Working Group comprised of cross-segmented representation, including: Faith, Addiction Recovery, Women’s Empowerment, Early Childhood Development and Parents, Sports & Recreation, Criminal Justice Reform, Homeless Assistance, and Small Business Owners and Economic Development
• Working Group workshops led by Instructors from Public Sentiment’s partners (NYU Department of Psychology and UC Berkeley) on social science research, data analysis/statistics, and grassroots organizing so that skills developed through this work remain within the community
• Predictive modeling to inform policymaker strategic communications and behavior change
• Automated sentiment analysis and other AI on voice-recorded response; potential to adapt cutting-edge emotional science detecting sentiment through facial expression captured on video-responses to survey
• Interactive data dashboard featuring raw and synthesized data, data visualizations, key findings, and additional resources, like additional social science research that helps contextualize community data
• Collaborative policy design workshops with residents and policymakers
• Sharing back data produced with respondents and community-based organizations for their own uses

Our approach centers on a resident-led approach. This means data collection is designed by Brownsville residents, analyzed and managed by them, and used for purposes they agree to. This process is effectuated with a Working Group comprised of Brownsville community residents who will help manage the process from start to finish.

Coordination with the community began with a comprehensive and strategic “funnel” outreach effort – starting with large state-wide organizations, and narrowing down to hyper-local organizations. This included working with the State’s Attorney General’s Office, City Council, City Agency Departments, and local community leaders, to identify interested partners. We contacted over 360 community-based organizations based or working in Brownsville. We then used our proprietary Community Mapping Matrix, which assesses over 30 segments of a community based on interest, demographic, and location, to build an inclusive and representative Working Group.

The Working Group is responsible for the following:

• Co-Designing the survey, including focus areas
• Distributing the survey through their networks via email, social media, and word-of-mouth
• Data analysis and development of key findings and “Narrative”
• Iterative policy design workshops with policymakers and decisionmakers

Our work helps local and state policymakers understand how residents within communities are being mentally, physically, economically, and socially effected by the pandemic, prioritize needs and policy solutions, and identify the best communication strategies and messages to help influence and change behavior. Together, these outcomes will help policymakers more effectively respond to the current pandemic, and help prevent similar outbreaks in the future. It will also help address inequities within communities so that lower-income communities are not disadvantaged in their responses to global pandemics. 

Oftentimes policy discussions involve policymakers and experts who define the technical details of a policy. While their perspectives are critical, equally important are the perspectives of the community who will live with the realities of any policy solution. 

Unfortunately, approaches used to involve people in can be:

Undemocratic: Groups of people may be excluded, creating gaps in data

Prescriptive: Pre-drafted surveys presuppose issues and solutions

Opaque: Information-flow between citizens and policymakers is largely unidirectional

Some organizations, like Bang the Table or Zen City, provide SaaS solutions. However, these “turnkey” solutions oftentimes leave policymakers with a lot of data, and not a lot of insight. Many municipalities have to hired data analysts to interpret the data, already creating gaps in the way insights are derived, and wholly removing residents from the process. Instead, our model creates a seamless process from start to finish, and integrates residents at every stage.

[PLEASE DO NOT POST THIS PORTION ONLINE]

In addition to our innovative model, our research teams from NYU and UC Berkeley are some of the foremost social scientists in their field. They are applying cutting-edge brain science to our research design and data analysis processes. We are exploring the potential of using a new coding-system that can classify 13 emotions automatically from video-recorded responses to survey questions. This would add depth and nuance to our quantitative data, and potentially allow for compelling stories and illustrative anecdotes to be shared with policymakers to contextualize policy recommendations.

Our model to genuine community engagement leverages electronic surveys, and a customized portal and user dashboard to collect and share back data and insights.

Within the portal, residents are able to take bi-weekly surveys, participate in Flanker tasks for additional research purposes, immediately access Covid-19 relief and recovery resources specific to their stated needs, see selected community-wide results from past surveys to understand context, and find other materials and resources related to the Covid-19 pandemic and its effect on communities and wellbeing. Proposed policy solutions will 

The survey is run through Qualtrics, and is housed within a custom-built website housed on NYU’s secured servers. Initial data visualization will be run with Python and R, and embedded in Jupyter Notebooks. Additional GIS visualization may be done depending on scope of data analysis.

In addition to these technological integrations, our model takes an innovative approach to iterative policy design – building a formal process through which residents and policymakers can address pandemic-related challenges and reform solutions incrementally, and over time. Policy design as a formal process has not been widely deployed in the United States; this is an opportunity to experiment with new formats of participatory government and civic participation as well.

Qualtrics is a vetted industry leader in electronic survey design and data management. 

Policy Design as a method of participatory government has been trialed successfully in several European countries including Finland (Universal Basic Income: https://99percentinvisible.org/episode/the-finnish-experiment/), and the United Kingdom (Policy Lab: https://openpolicy.blog.gov.uk/category/policy-lab/). In the United States, New York City’s DOT has started to explore policy design with its Select Bus Service, currently underway, and showing positive results (SBS: https://ny.curbed.com/2019/10/18/20919729/new-york-transportation-bus-lane-transit-priority)

We believe that policymakers and communities can develop smarter and more effective policies that adequately address residents’ needs with better data and sustained and iterative policy development. Integral to understanding how residents are faring from the pandemic is the development of a robust survey tool, that identifies key stressors posed by the pandemic, and the mental and physical impacts those stressors have had on people’s lives. To do this, we have designed a survey that achieves the following key objectives:  

• Understand what stressors Brownsville residents are facing.
  
  • Which stressors are Brownsville residents reporting the most/least? 
  • How do these stressors relate to psychological health? 
•  Provide Brownsville residents with resources for better ways of coping with stress and regulating negative emotions. 
  • How are residents coping with stress?
  • How are residents currently regulating their negative emotions
• Understand the physical and mental health of Brownsville residents.
• Understand whether Brownsville residents have a sense of social connection.
• Understand what sheltering in place looks like for Brownsville residents.
• Understand best strategies for effective communications to the community, including specific messengers and channels.
  • Which sources of technology do Brownsville residents use and what do they use them for?
  • Which messengers do Brownsville residents trust considering groups and people like Community Board 16 to President Trump.
• Predict changes of residents’ behaviors, outlooks, or responses over time.
• How do the questions we asked above change over time?
• Do mean levels of stress change over time?
• How does the relative importance of certain stressors change over time?
• Understand how concerns among residents differ between demographics 

Through rigorous, science-based synthesis, we will then be able to share insights, policy recommendations, and communications strategies that most effectively help policymakers meet the needs of the community. 

Directly, we will be providing skills development training to 15 residents, and providing at minimum 1,000 residents connections to direct relief and recovery assistance programs. Resulting policy solutions would likely be focused on the entire community of Brownsville, Brooklyn, which has a population of 112,000.   Within the next year, we will be working City-wide, potentially with 5,000 residents, with policy solutions impacting 500,000 residents. In 5 years, we aim to be integrated within New York City, and expanding to other cities in the country, and around the world – especially as second (or third) waves of the pandemic occur. 

Within the next year, we will be continuing our work in Brownsville, potentially expanding the social and economic impacts intersecting with the Covid-19 pandemic. Within the next five years, we will be helping local policymakers and communities proactively build more dynamic and resilient social engagement structures that can be invaluable in responding quickly and effectively to global pandemics, especially in areas where access to information is low. 

Due to the social nature of our work, our process may be influenced in both positive or negative ways:

- 2nd wave of Covid-19 outbreak – potentially effecting community engagement, survey distribution, and community conversations (While this would impede our ability to gather the Working Group in-person, and conduct robust grassroots organizing effort in-person, we have built this process for remote participation (including an electronic survey, and video-conferencing meetings, that would allow us to adapt easily back to current conditions.

- Social unrest – Brownsville is heavily involved in current protests around policy brutality. While aspects of our survey touch on this issue, it does so through the lens of Covid-19. On the other hand, if protests increase the spread of Covid-19, this may once again be top of mind for most residents -   As New York eases social distancing policies and “re-opens,” there will likely be more City-wide data accessible for analysis and benching marking purposes, against which we can evaluate the impacts within Brownsville. This additional information will help illustrate stark contrasts between NYC neighborhoods that can inform and prioritize public policy.

We have designed our process to work within a 100%-remote context - holding working groups on video conference calls, distributing the survey electronically, and coordinating the volunteers who are able to administer the survey via phone for residents who do not have consistent internet connection. Should a second wave strike the City, we would be flexible and adaptive in our response, uses the same existing tools we have in place from the initial wave.

Additionally, for any social unrest that may ensue due to upcoming elections or changes (or lack thereof), we will again deploy effective adaptations. 

N/A

We have a team of 18 doing this work between our organization and our strategic partners. This includes 1 F/T, 1 P/T from Public Sentiment, 6 part-time from UC Berkeley, 5 from NYU Psychology, 1 P/T from NYC/Grid, and 4 from NYU-CUSP (Center for Urban Science and Progress).

Our multidisciplinary team is comprised of 50/50 women and men from around the world, specializing in cognitive science, engineering, design, public policy, grassroots organizing, data analysis, and data visualization. We span ages 25 - 50, and have vast experience working with marginalized and vulnerable populations, including immigrants, veterans, low-income families, and communities of color. We have strong alignment on the project objective of improving people's qualities of life by reducing the social, economic, and health challenges posed by public health crises, and have strong institutional backing and resources that can help us sustain our work.

We are working with the following organizations:

NYU-CUSP - City-wide data analysis, visualization, and benchmarking

NYU Department of Psychology (mindHive) - Survey development, platform and dashboard design

UC Berkeley (Greater Good Science Center) - Survey development, data analysis, Working Group Workshop Instruction 

NYC/EDC's Urban Tech & Grid Network - Coordinating policymaker participation 

We are a 501c3 nonprofit; we accept grant money and donations. We also are a fee-for-service consultant to municipalities and other civil-society organizations.

To help minimize costs, we partner with labs at academic institutions, who are able to apply their institutional resources and funding to our work to help further their own research aims. 

We have secured funding from one institutional backer to date for our current work, and are in final discussion for additional grants with two other foundations, totaling $48,000.

We are also exploring major long-term grants through the NIH and NSF.

Public Sentiment sees great promise and potential in joining the Solve community for two reasons. First, Solve offers an opportunity to draw on the vast technical expertise of MIT’s faculty and affiliated practitioners. Because Public Sentiment seeks to leverage new technologies, like natural language processing for multimedia – an area of focus at the Media Lab, there is great possibility. Second, Solve brings together action-oriented, socially-conscious entrepreneurs from around the world. With ambitions to one day work globally on community engagement issues related to social and environmental justice issues, we would greatly benefit from inclusion in such a diverse and active network.  

We aim to build more advanced automated sentiment analyses tools, especially fit for multimedia (specially, video), to help scale and accelerate analyses. Moving away from text-base analysis will make data collection more inclusive and accessible to people of all ages, language, and digital literacy levels, and help scientists gain a more nuanced and accurate reading of how people are feeling.

Additionally, we are looking to bring talented and diverse talent into our organization, ensuring that we are incorporating global perspectives and experiences in our service design. Working within the Solve network would help expose us to a broad talent pool of technologists and similarly-minded problem solvers.

Finally, the Solve platform would help us reach target audiences of researchers, technologists, and potentially policymakers who might be interested in our service model. 

Public Sentiment would like to learn from, collaborate with, and help increase the impact of, other Solver Teams, including: LifeBank, to coordinate essential medicines and PPE to low-income communities like the ones where we work; Design Thinkers Without Borders (helping inform curriculum design, and expedite designers research process, by automating aspects of the data collection and analysis process), and; Sustainable Community Development at Standing Rock (supporting the Solver Team’s effort to identify what community needs, concerns, and ideas exist to help guide the project’s focus and development).

Public Sentiment is a community engagement organization that uses data and dialogue to help policymakers and communities build smarter policy solutions.

We believe that when people have access to trusted information, they can increase their understanding about themselves, others, and their place in the world. 

For effective Covid-19 Recovery, we are helping residents and decisionmakers develop effective strategies for immediate pandemic relief, and long-term recovery. 

Our approach involves data analysis on existing City-wide data, a year-long, bi-lingual, bi-weekly survey to track residents’ evolving needs, resident-led data analyses, and iterative, collaborative policy design with policymakers. 

This work:

1) Immediately connects residents to appropriate relief organizations and government agencies, based on their specific challenges;

2) Informs policymakers, researchers, and service providers about the mental, physical, and behavioral effects of a pandemic to inform future prevention, relief, and recovery strategies and identify best communication strategies and messages to influence and change behavior. 

We address intersecting issues like poverty, education, and gender to help address inequities within lower-income communities so that they're not disadvantaged in their responses to global pandemics. 

In addition, we also aim to build strong social fabric and greater understanding and trust in public institutions and businesses within the community and between residents, nonprofits, and policymakers.

Funding through Elevate would allow us to hire additional staff, specifically within our Communications Team, to help produce research briefs for policymaker audiences, as well as general public, and alert decisionmakers about our new and effective model for civic participation and policy design. 

Like The Patrick J. McGovern Foundation, Public Sentiment believes in the powerful combination of human and artificial intelligence to improve the quality of life on Earth. Public Sentiment seeks to integrate underrepresented citizens into the policymaking processes that will shape their lives, communities, and countries. It will do this using AI technologies to improve information-gathering efficiencies and quality, and groundbreaking research in brain and behavioral sciences to facilitate dialogue and understanding.

Public Sentiment’s approach relies on AI technologies – analyzing citizen-submitted audio and video stories to identify critical hopes and needs. This data provides rich nuances – like intonation or facial expression – that can be used to inform policy recommendations. However, current sentiment analysis technologies work best on text-based data. This means Public Sentiment’s multimedia must first be transcribed for analysis, and that the full value of citizens’ content is unrealized.

Leveraging our team of skilled technologists, neuroscientists, and researchers, Public Sentiment will use the Foundation’s support to refine and augment the text-based AI technologies currently available to take fuller advantage of data-rich multimedia content. With higher-quality insights, Public Sentiment will be better able to help policymakers craft policies responsive to, and reflective of, citizens’ needs and aspirations.

Public Sentiment is a community engagement organization that uses data and dialogue to help policymakers and communities build smarter policy solutions that meet everyone's needs.

We believe that when people have access to trusted information, they can increase their understanding about others, themselves, and their place in the world. 

We aim to build strong social fabric within the communities - between residents, nonprofits, and community leaders and elected officials - by increasing communication and data-sharing to increase understanding and trust in the public institutions and businesses within the community and help ensure better working relationships in the future. This is in the spirit of SDG Goal #16. Within the context of our current project around Covid-19 response and recovery in New York City, this also helps advance SDG Goal #3.

In communities everywhere, the Covid-19 pandemic has created new and urgent challenges, and exacerbated deeply-entrenched challenges long in need of systemic reform. It has highlighted disparities in access to quality healthcare, reliable information, and use of effective preventative practices. These disparities fall largely along racial and socio-economic lines.

Our approach centers on a resident-led approach. This means data collection is designed by community residents, analyzed and managed by them, and used for purposes they agree to. 

We believe this model of inclusive participatory government can be applied to context and geographies around the world, and therefore, helpful for accomplishing the SDGs by 2030.

With additional funding from The People's Prize, Public Sentiment would be able to approach lower-resource governments to assist them with their participatory government efforts at greatly reduced rates (or pro-bono).