We’re excited to announce that DataKind has launched a cohort of seven U.S.-based, economic empowerment-focused DataCorps® projects! This cohort of projects is part of a portfolio of work that we kicked off in 2021 to leverage data science and AI with mission-driven organizations focused on economic resilience across the globe.
With support from Google.org, DataKind has partnered with seven social sector partners to apply data science and AI in four key thematic areas within economic empowerment: financial inclusion, post-secondary education, safe and affordable housing, and poverty alleviation.
DataKind staff and volunteer scopers developed projects in direct collaboration with our partners to ensure adoption and sustainability of the final products. We recruited 55 expert, pro bono data scientists to launch our DataCorps® engagements. The teams, with specifically selected technical skills to support each project need, are working with partners over the next six months. The end result? Fully co-created solutions for our partners so that they can continue to extend their reach and maximize their impact in economic resilience and empowerment.
This work demonstrates DataKind’s value add in the economic empowerment space and highlights how data science and AI can be a catalytic tool to extending an organization’s reach and impact at different levels of a system. As the DataCorps® teams continue to build their solutions, DataKind aims to identify which of these solutions could be scaled, driving impact that extends beyond individual organizations.
Without further ado, meet our seven project partners below, and learn how we’re working together to advance economic empowerment!
Women in the US receive only 4% of all small business loans from mainstream financial institutions. Grameen America (GA) is dedicated to helping entrepreneurial women who live in poverty build businesses to enable financial mobility through microloans, financial training, and direct support to members. GA measures their performance by the number of active loans and active members in each relationship managers’ portfolio. Currently, each step of GA’s multi-step process for viewing this data equates to hours of people’s time; on average, staff take two hours to complete the process for each report with an average of 10 reports per week. DataKind is partnering with GA to develop carefully curated data visualization dashboards to complement and augment current daily, weekly, and monthly reports. Our project aims to automate the creation and visualization of these reports by transforming text based data into charts, graphs, and maps using business intelligence tools. This can provide managers with the ability to identify trends and clients most in need of support. This, coupled with predictive analytics around loan default and attendance, will provide the GA team with the ability to improve retention and loan accountability so that their clients are able to continue their financial and professional growth and build their economic resiliency. This solution also provides GA staff with the time required to further scale their operations beyond their current 150,779 clients with a specific focus on Black women entrepreneurs.
Across the US, less than half of small businesses have access to the financial services and capital they need in order to meet their credit needs. Yet business owners and entrepreneurs need access to a variety of credit sources in order to start, maintain, and grow their businesses. Kiva is on a mission to expand financial access to underserved communities through crowdfunding loans and unlocking capital for the underserved, improving the quality and cost of financial services and addressing the underlying barriers to financial access around the world. DataKind is partnering with Kiva US to understand which customer characteristics correlate with improved business performance, household income, customer satisfaction, and higher repayment rates, so that their US lending program can approve loans to those who are not only underserved by the formal financial sector but are also most likely to benefit from Kiva loans.
Additionally, DataKind and Kiva will use natural language processing models on their borrower profile stories to determine the characteristics of a successful story such that an entrepreneur is able to achieve their full business loan request amount. This information will allow the Kiva team to further strengthen borrower profile stories so that a greater number of borrowers on their website can receive their full loan amount, so that small business owners across the US are able to invest in their professional expansion.
Students at universities across the US have expressed concerns about their ability to advance and complete their degrees within the allotted time without having clear knowledge as to the courses they need when they need them most. Complete College America seeks to tackle this challenge with a data-driven approach. DataKind is partnering with Complete College America and one of their direct implementation partners, San Antonio College, to create and test a course scheduling tool that will help alleviate a known pain point around course accessibility: providing classes driven by student demand rather than teacher supply. Adapting the methodology from a course enrollment prediction publication about a predictive model built to forecast course enrollment figures at Harvard University, DataKind and Complete College America will build tooling to help resource-constrained academic programs create schedules that more accurately reflect student demands. This solution has the potential for direct impact on San Antonio College’s approximately 20,000 students, and a strong scale potential through Complete College America’s alliance of 48 states, systems, and consortia, and the hundreds of institutions represented by that network.
Across the US, four-year graduation rates for transfer students who move from community college to a senior (four-year) institution are low: only 42% of transfer students obtain their Bachelors’ degree within six years of starting at a community college, compared to the national average of 60% for non-transfer students over the same time period. In order to address the contributing factors, DataKind will team up with John Jay College once again and collaborate on a complementary solution to the co-created 2019 CUSP graduation tool that has helped an additional 3,200 students complete their college journeys. Recognizing that transfer students face a host of challenges in reaching graduation, this tool will aim to help academic counselors identify and direct support to transfer students who have defined risk characteristics, with hopes of ensuring their academic success. Using historical student data, the model will leverage machine learning algorithms to generate the probability of a transfer student not completing their degree within six years, providing recommendations on when this risk is highest, so that John Jay College financial and academic advisors can provide students with the necessary support to help them graduate from their baccalaureate programs within the prescribed six years.
Every year, 5 million Americans lose their homes through evictions and mortgage foreclosures. This forced displacement is intensely traumatic and has been linked to poor outcomes, including adverse health impacts, gaps in educational attainment, and chronic homelessness. Building on the collaboration during the Economic Resilience Discovery Day, DataKind will work with Bright Community Trust and their Central Florida regional housing trust, housd, to provide leaders in the four Central Florida counties of Orange, Seminole, Osceola, and Lake with actionable information, such as trends in evictions and where federal emergency rental assistance was distributed, to increase housing security by bringing together previously siloed data streams and using a combination of statistical, geospatial, and predictive modeling techniques. This partnership will allow housd to be data-driven and data-responsive in their reaction to community needs related to safe and affordable housing. The direct impact of this partnership will be on the residents of these four counties (12.3% of the Florida population) to increase access to safe and affordable housing as provided by the nearly 100 organizations and individuals represented in housd’s Evictions and Foreclosures Group.
This work will be completed through the expansion of the digital public good toolkit – the Foreclosure and Eviction Analysis Tool (FEAT) – that DataKind and New America’s Future of Land and Housing Program created in 2021 and the learnings regarding housing loss developed through the Displaced in the Sunbelt research program (which included contributions from Bright Community Trust). Key outputs will span three work streams: identifying housing insecure areas within the target geography, assessing the emergency rental assistance program, and predicting housing loss and targeting community aid.
The Center on Rural Innovation (CORI) is dedicated to closing the rural opportunity gap by advancing economic prosperity through the creation of inclusive digital economy ecosystems that empower rural Americans to create tech jobs and launch scalable tech companies. One of CORI's initiatives is the Rural Innovation Initiative, which works with rural communities to create strong digital economies and help them chart their own courses toward prosperity.
To determine rural communities that could benefit from CORI’s initiative, the team identifies and assesses each community’s current assets, opportunities, gaps, and challenges to identify where they could provide direct technical assistance for the development of communities’ digital economy and ecosystem. Currently, the CORI team is filtering data across multiple sources to identify communities that would most benefit from their support. The development of a robust weighting system will provide further granularity around which communities to target, utilizing a machine learning algorithm that will draw upon known successful community digital economy scale development.
DataKind is partnering with CORI on a machine learning driven quantitative decision support tool with specific weights for categories most relevant to their inclusion criteria (e.g broadband fiber and access to technology jobs) so the team can have a more sophisticated selection method to guide direct outreach to rural communities in the US. This tool will better predict the likelihood of successful outcomes amongst communities CORI identifies in its selection process, helping those with the highest potential and need to benefit from technical assistance to create and strengthen their digital economies.
Every day, millions of Americans rely on social safety net services to build their economic stability such as access to unemployment benefits, food banks and food pantries, and safe and affordable housing options. While COVID-19 highlighted the importance of accessing these services for the stability of individuals and families, it also demonstrated the large gap in the timely provision of these services to those most in need. DataKind will work with United Way Worldwide to develop a demand forecasting model through time series forecasting and regression modeling to catalyze the power of local data, collected by local United Way (UW) chapters, their nonprofit partners, and records of 211 calls, to provide a data-driven assessment of the situation in each community on key health, education, and financial indicators, all with an equity lens. This information will provide DataKind, local UW chapters, and UWW the ability to determine where services are needed most for the subsequent quarter, and where they are currently being distributed. This partnership will be piloted with the United Way of South Central Michigan’s (previously United Way of Jackson County) and its Jackson County Care Hub, a comprehensive system that assesses incoming clients for issues related to the Social Determinants of Health, with the opportunity to then be replicated, transferred, and scaled across the 1,000+ United Ways across the country.
Header image courtesy of Kiva / Brandon Smith.
As always, thank you for your support of this critical work!