Nava Home
Public Benefit Report2024

Influencing ecosystems

Changing the fabric of service delivery requires collaboration between government and technologists. Our interdisciplinary expertise guides us toward systems that are flexible, resilient, and human-centered.

Illustration of seven people inside a city hall. In the foreground, a Black nonbinary person with short, curly hair holds a cup of coffee. To the right, a middle-aged Latinx woman talks to a young Indigenous woman and a young Black man with a prosthetic left arm. The Indigenous woman and Black man sit in chairs. In the background, a young Asian man holds his toddler.

  • Ensuring ethical experimentation with AI in the public sector

    Artificial intelligence (AI) has the potential to demystify the public benefits application process for applicants and for professional benefits navigators like case workers and community outreach specialists. However, teams must deploy AI intentionally and ethically to avoid causing harm to vulnerable populations. In many cases, this means using AI to augment what navigators do, not using it to eliminate the human interactions that take place between navigators and the public.

    Development teams can mitigate the risks associated with AI by building systems alongside those who will use the technology — often called human-centered design. We also recommend testing new AI tools in low-risk environments, defining success metrics and evaluating, iterating, and rinsing and repeating — also called agile development. This is consistent with some of the directives put forth by the Biden-Harris administration’s executive order on AI.

    We’ve championed these practices since our founding. Now, we’re applying them to our project creating, testing, and piloting an AI-powered tool that may help government agency staff and benefits navigators easily identify which families can enroll in key public benefit programs. 

    For this work, we’re exploring how human-centered design, agile development, and rigorous evaluation might enable us to responsibly implement AI models while improving outcomes for the public. In addition to providing immediate value for benefits navigators, we hope this project’s insights will inform AI policy and implementation.

    To learn more, read our op-ed, originally published in Tech Policy Press. 

    9

    U.S. Department of Agriculture, Food and Nutrition Service, Reaching Those in Need: Estimates of State SNAP Participation Rates in 2020

    10

    U.S. Department of Agriculture, Food and Nutrition Service, National and State Level Estimates of WIC Eligibility and Program Reach in 2021

    Image caption

    Human-centered design, agile development, and rigorous evaluation can help teams build AI-powered tools that help vulnerable populations rather than hurting them.

    Illustration of a Black nonbinary person with short, curly hair in an office. They look at a chatbot service on their smartphone.
    Illustration of seven people inside a city hall. In the foreground, a Black nonbinary person with short, curly hair holds a cup of coffee. To the right, a middle-aged Latinx woman talks to a young Indigenous woman and a young Black man with a prosthetic left arm. The Indigenous woman and Black man sit in chairs. In the background, a young Asian man holds his toddler.

  • Conducting user research for human-centered experimentation with AI

    We’re creating, testing, and piloting an AI-powered tool that may help government agency staff and benefits navigators — or professionals who help people apply for benefits — easily identify which families can enroll in key public benefits programs. During the first phase of this project, we partnered with Benefits Data Trust to conduct user research with stakeholders at every level of the benefits application process.

    We spoke with benefits navigators, program beneficiaries, and strategic advisors from organizations in the public benefits space. Our research helped us identify core needs that beneficiaries and navigators have throughout the benefits application process. If those needs aren’t being met, we considered it a potential use case for an AI-powered tool. 

    For four of the use cases, we designed human-centered experiments to gauge whether AI models might support people’s unique needs. Our experiments helped us identify one promising use case — building a chatbot to help answer navigators’ questions — which we’ll pilot with navigators.

    Early on in the project, we decided that all of our experiments would keep a human in the loop, meaning that navigators vet a tool’s output to ensure accuracy and reduce the risk of harm. 

    To learn more, you can read our case study on our user research and our case study on human-centered experimentation.

    Image caption

    Nava Labs is testing AI-powered tools that might help benefits navigators connect families to critical resources.

    Illustration of a young Indigenous woman sitting at an office desk and talking on her cell phone. There is a laptop, monitor, and notebooks on the desk in front of her. Someone working on a computer sits across from her.
    Illustration of seven people inside a city hall. In the foreground, a Black nonbinary person with short, curly hair holds a cup of coffee. To the right, a middle-aged Latinx woman talks to a young Indigenous woman and a young Black man with a prosthetic left arm. The Indigenous woman and Black man sit in chairs. In the background, a young Asian man holds his toddler.

  • Communicating the successes of state unemployment insurance programs

    The pandemic strained state unemployment insurance (UI) programs and laid bare many long-standing issues with our country’s UI system. As a result, the U.S. Department of Labor (the Department) received significant resources from the American Rescue Plan Act (ARPA) to collaborate with states on technology modernization projects. Nearly three years into this effort, Nava helped the Department publish a report on how states are using ARPA funds. 

    “Insights and Successes: American Rescue Plan Act Investments in Unemployment Insurance Modernization” outlines how ARPA funding is helping states detect and prevent fraud, promote equitable access to UI, and ensure timely payment of benefits. Nava helped with researching, editing, designing, and publishing the report and accompanying blog posts on the Department’s reference site. 

    We also worked with the Department to send the report to over 600 people from roughly 145 organizations in the UI space, and we provided event support for the Department’s congressional briefing webinar. Over 200 people attended the congressional briefing, demonstrating strong interest among decision-makers in modernizing UI. 

    Telling the story of UI modernization and why this funding matters helps the Department share out the great work they’ve put into improving critical UI programs across the country. This project also demonstrates Nava’s ability to partner with a government agency to execute several strategic communications initiatives. 

    To learn more, read our case study. 

    Image caption

    We helped the Department communicate state UI successes, helping build momentum for modern UI programs that are resilient, equitable, and adaptable.

    Illustration of a middle-aged Latinx woman standing at a podium. There is a U.S. flag behind her.
    Illustration of seven people inside a city hall. In the foreground, a Black nonbinary person with short, curly hair holds a cup of coffee. To the right, a middle-aged Latinx woman talks to a young Indigenous woman and a young Black man with a prosthetic left arm. The Indigenous woman and Black man sit in chairs. In the background, a young Asian man holds his toddler.

  • Leveraging user research to improve unemployment insurance delivery

    After the pandemic exposed weaknesses within our country’s unemployment insurance (UI) technology, the U.S. Department of Labor (the Department) and state UI agencies received a historic $1 billion in American Rescue Plan Act (ARPA) funding to modernize state UI programs. As part of this work, the Department partnered with Nava to spearhead qualitative research that resulted in targeted recommendations for how states can improve UI delivery. 

    We helped the Department conduct extensive user research with claimants, state workforce agencies, UI experts, and community-based claimant advocates. Our research helped us identify recommendations for improving the initial UI application and weekly certification. We then turned our recommendations into prototypes and tested them with UI claimants and applicants. Finally, we published our learnings in several articles on the Department’s reference site so states can leverage them when modernizing their UI applications. 

    To learn more, read our case study. 

    11

    U.S. Department of Labor, Employment and Training Administration, Unemployment Insurance Modernization, Customer Experience

    Image caption

    Qualitative research enabled us to identify and publish targeted recommendations that states can use when modernizing their UI programs.

    Illustration of a young Black man with a prosthetic left arm. He sits at a table and talks to a group of three people. There are charts and notes on the table and around the room.
    Illustration of seven people inside a city hall. In the foreground, a Black nonbinary person with short, curly hair holds a cup of coffee. To the right, a middle-aged Latinx woman talks to a young Indigenous woman and a young Black man with a prosthetic left arm. The Indigenous woman and Black man sit in chairs. In the background, a young Asian man holds his toddler.

  • Reducing the burden of applying for public benefits

    Facing a Financial Shock is a Biden-Harris administration Life Experience project aimed at connecting people experiencing financial distress to available public services. Two of the project’s collaborating agencies — the Centers for Medicare & Medicaid Services (CMS) and the United States Digital Service (USDS) — are partnering with Nava to increase access to programs like Medicaid and the Supplemental Nutrition Assistance Program (SNAP) by streamlining the income verification process.

    Together, we aim to help agencies verify applicants’ income eligibility through existing data sources rather than requiring applicants to submit income information. This idea is called Income Verification as a Service, and it’s something we’ve championed since 2016, when we worked with CMS on HealthCare.gov’s income eligibility determination services. 

    To ensure our work is human-centered, we’re practicing several forms of participatory design. This has informed which solutions we pursue, such as a plain-language playbook that state agencies can use to simplify content around income verification. We’ve also helped prototype a tool that gives applicants the option to automatically share their payroll data with benefit programs. Lastly, to help states more easily determine income eligibility, we’re working on a solution that connects states to the right data sources at the right time so that the onus to provide income data is not on applicants. 

    To learn more, read our case study. 

    12

    Urban Institute, "A Safety Net with 100 Percent Participation: How Much Would Benefits Increase and Poverty Decline?"

    Image caption

    Easing the income verification process can help more eligible people access important services like food benefits.

    Illustration of a young Asian man holding his toddler at a grocery store. The man reaches for a jar on a shelf.
    Illustration of seven people inside a city hall. In the foreground, a Black nonbinary person with short, curly hair holds a cup of coffee. To the right, a middle-aged Latinx woman talks to a young Indigenous woman and a young Black man with a prosthetic left arm. The Indigenous woman and Black man sit in chairs. In the background, a young Asian man holds his toddler.
Next Section
Previous Section