How Integrated Data Can Support COVID-19 Crisis Response and Recovery
By Emily Berkowitz & Matthew Katz
This blog was updated on November 18, 2020. It is the first of a two-part series on administrative data reuse for crisis response and recovery. Read part 2 here: https://aisp.medium.com/established-data-infrastructure-enables-local-governments-to-meet-need-during-natural-disasters-a488fbb4a409
As COVID-19 continues to create life-changing impacts, interventions that respond to the increasingly complex and interconnected needs of individuals, families, and communities are acutely needed. Effects of the crisis cut across services and sectors, disproportionately impacting Black, Indigenous, and People of Color (BIPOC)[1] communities and thwarting progress toward more equitable economic mobility.[2]
COVID-19 has revealed long-time structural inequities already faced by BIPOC in the U.S. and surfaced new problems. For example, Black workers are employed disproportionately in front-line positions (e.g., store clerks, child care providers) that increase the risk of contracting and spreading coronavirus. Likewise, issues such as lack of health insurance benefits, higher rates of living in dense or crowded housing, and persistent racial wage gaps leave many families of color with few cash reserves to draw on in the case of job loss.
State and local leaders are called upon to respond to the immediate harms of COVID-19. Yet, with a looming recession threatening to undo gains among marginalized groups — particularly the Black middle class — tools to understand long-term impacts on economic well-being are also urgently needed.
Administrative data[3] — the information collected during the course of routine service delivery, provide an essential tool to help understand short- and long-term impacts of the pandemic. Several jurisdictions now have the capacity to link administrative data across programs in order to better understand how individuals interact with multiple systems. As the COVID-19 crisis reveals weaknesses in the U.S. social safety net, communities with integrated administrative data infrastructure can use this capacity to identify vulnerable populations and unmet needs . For example, youth who “age out” of the child welfare system or individuals who experience chronic homelessness often remain invisible when using traditional methods or administrative records from a single source.
This blogpost demonstrates how nimble state and local data integration efforts have leveraged their capacity to quickly respond to and understand the impacts of COVID-19.
What are integrated data?
Integrated data help governments respond more quickly to crises by enabling cross-agency collaboration and streamlining bureaucratic processes. When data are linked they can provide unique insights into how programs impact people across domains and over long periods of time. At AISP, we typically refer to these efforts as integrated data systems (IDS), but they are sometimes known as data collaboratives, or data hubs. IDS are typically built for routine use, establishing secure infrastructure for data access and strong protocols to govern how data are used and protected. While building and scaling an IDS can be time-intensive, the benefits are considerable — especially during crises. By breaking down the silos in which data are held, jurisdictions can foster stronger partnerships between agencies and leverage insights to adjust policies to reflect community need.
Integrated data provide an excellent avenue for the holistic study of economic mobility that extends beyond traditional indicators of economic success. An IDS can be used to identify root causes of mobility barriers as well as the ripple effects of hardship, such as housing segregation and lack of access to quality education. Data can also be linked across an individual’s lifespan, providing insights into key life course transitions, such as kindergarten entry, as well as transitions into or out of a program, or institutional setting. Increasingly, sites are developing the ability to accurately match members of the same household to consider the multigenerational impacts of programs.
Regardless of the reason for data integration, sites must grapple with key questions and concerns related to equity and surveillance due to the inherent risks (and benefits) of administrative data reuse.
See A Toolkit for Centering Racial Equity Throughout Data Integration for a nuanced discussion of risk vs. benefit in data sharing and for actions your site can take to center racial equity in your work.
How are integrated data being used in response to the COVID-19 crisis?
States and counties with IDS capacity are already working to understand which policies and can mitigate the predicted negative outcomes of COVID-19. The economic and social impacts of the novel coronavirus have revealed not only the underlying weaknesses of the social safety net but also the ability of certain efforts to pivot operations in response to crisis. Consider state-level efforts in Ohio and California that, at the start of state-wide lockdowns, were able to quickly leverage their IDS to support demand for front-line employment and child care for essential workers.
In Ohio, the InnovateOhio Platform (IOP) was leveraged to build an interactive website to link individuals whose employment was impacted by COVID-19 to employers with current job openings. Hosted by the Department of Administrative Services Office of Information Technology, IOP demonstrates the benefit of having a state-level agency dedicated to the governance and use of administrative data. The state also recently released the COVID-19 Ohio Minority Health Strike Force Blueprint to guide their “comprehensive and systemic approach” to address the virus’ disparate impact on Ohioans of color, including indicators to track progress on stated goals.
In California, the California Health and Human Services Agency (CHHS) and USC Children’s Data Network built an online tool, called mychildcare. After the temporary closure of most daycare centers in the state, mychildcare was designed to connect essential workers to child care providers with open slots. In collaboration with the Governor’s Office, CDN and CHHS helped build the searchable, web-based interface in 10 days. Mychildcare was successful in visualizing, for the first time, the availability of child care slots statewide. It also served as a validation of the Research Data Hub, a secure, cloud-based research enclave for hosting linked research data sets, which was at that time still under development. Mychildcare data, now hosted within the Research Data Hub, are also being used to answer requests for data from federal and state partners to improve planning (e.g., the number of facilities open now that weren’t open in the prior month, the number of child care slots being requested).
These operational use cases demonstrate how integrated data can keep families and businesses afloat by facilitating employment opportunities and assisting essential workers — all during a national crisis.
Similarly, the Linked Information Network of Colorado (LINC), a partnership between the State of Colorado and the University of Denver’s Colorado Evaluation and Action Lab, were able to use data from an ongoing LINC project to help connect early educators to childcare and preschool providers in need of staff. Prior to the pandemic, LINC was using their IDS to analyze the state’s early care and education (ECE) workforce. After learning about Early Milestones Colorado’s work to fill temporary gaps in the ECE workforce due to COVID-19, analysts were able to repurpose their datasets to generate a list of ECE professionals who were ready to work. LINC’s partners then used the list to reach out to over 30,000 ECE workforce members to ensure essential workers had reliable childcare. This example underscores that much of data integration work is relational, rather than technical and demonstrates what’s possible with partnerships among universities, public, and non-profit agencies.
Moving forward, LINC is using integrated data to understand the pandemic’s impact on the ECE workforce. A recently launched dashboard surfaces gaps in recruitment, retention, and professional development of ECE professionals. The dashboard will be updated during and after the pandemic and aims to help the state be responsive to changing conditions for the ECE workforce.
Numerous additional examples from IDS sites across the U.S. that are using linked administrative data in response to the COVID-19 crisis can be found on the AISP website.
How can integrated data help long-term COVID-19 recovery?
In the long-term, IDS are a powerful tool to help identify opportunities for improving the social safety net, to tell us which policies are beneficial, and to help quantify the ways in which racism — both systemic and overt — continues to harm the BIPOC community . In order to manage public health disasters and their disparate impacts in the future, more communities must develop the ability to ethically link data across systems in ways that encourage an ongoing, evaluative process and incorporate community feedback. Data are not a solution in and of themselves. Rather, sites using cross-sector data should be rooted in a mission that is backed by established governance processes designed to ensure the work is done with public trust and security protections in place.
As we consider the value of IDS, it bears repeating that the need and loss in the wake of COVID-19’s spread is exposing existing hardships rather than uncovering new problems created by the virus. Our response must reflect continued attention to how historical legacies of harm (e.g., redlining, school segregation, Indian Child Welfare Act) continue to impact BIPOC across a variety of domains.
The interactive COVID-19 Vulnerable Communities Data Tool, for example, was built by Public Health Seattle & King County and Communities Count to identify areas where social conditions have caused increased risk of contracting the virus that make social distancing more challenging. Data show areas in King County that may be more vulnerable to COVID-19, making it easier to direct resources to communities in need. While this effort does not leverage client-level data held by the county’s IDS[4], it highlights the value of a data-driven approach to crisis response. By identifying key areas of dispossession driven by racism, the Vulnerable Communities Data Tool sets a foundation to center racial equity and marginalized groups.
What does COVID-19 mean for integrated data?
As the pandemic upends common standards in the field, the usability and reliability of data models is changing. Child welfare reports are at perplexing lows — below the typical dip that occurs in summer months — suggesting that traditional measures of abuse and neglect may not be as reliable in the current moment). School attendance and test scores are no longer relevant to understanding the influence of early childhood and education as they once were because in-person instruction has stalled or shut down. Similarly, metrics like well-child health visits or afterschool program participation have dramatically changed due to lockdowns. Yet, SNAP benefits are being leveraged flexibly and extensively in response to the pandemic and offer an important measure for future work.
Lockdown and remote work also reveal limitations of certain privacy and security measures at a time where limited contact is lifesaving. For example, efforts where data can only be accessed using on-site servers saw work put on pause indefinitely when lockdown measures went into effect. Other efforts were stymied by legal agreements that do not allow data integration for operational use, limiting how information can be used in real-time. As sites work to modernize their crisis response capacity, they may also need to amend their legal agreements to alter how data are stored, shared, and used in practice. On the other hand, some sites are reporting that data which were difficult to access prior to the start of the pandemic, such as Medicaid and public health, are now being leveraged more rapidly to respond to the crisis. Anecdotal reports suggest data sharing agreements are being executed more quickly, and that calls for more access to administrative information are only increasing.
With demand for data and awareness of inequities both at an all-time high, the risks associated with data reuse are front and center and our field must meet them head on. But these challenging times also create new opportunities for us to better utilize integrated data for social good and fundamentally rethink the social safety net with equity in mind.
At AISP, we are working hard to document and embrace these opportunities in collaboration with communities. In our next blog post, we will explore the value of integrated data in responding to climate disasters and provide a series of examples that demonstrate how communities with data integration capacity are adapting to meet new challenges with an intentional equity focus. Read part 2 here <link>.
For more about how AISP Network members are responding to the evolving crisis, please visit our website and subscribe to our monthly newsletter.
[1] AISP intentionally uses the acronym BIPOC (Black, Indigenous, and People of Color) in our publications as a term that seeks to recognize the unique experience of Black and Indigenous People within the United States. We recognize that naming is power, and we remain committed to using language that supports pro-Blackness and Native visibility while dismantling white supremacy. Read more here: https://www.vox.com/2020/6/30/21300294/bipoc-what-does-it-mean-critical-race-linguistics-jonathan-rosa-deandra-miles-herculesn
[2] Our understanding of economic mobility and the many factors that affect individuals, families, and communities across the life course aligns with the Urban Institute in Boosting Upward Mobility: Metrics to Inform Local Action (2020). The authors approach economic mobility holistically, moving beyond traditional measures of economic success and assessing other social and community factors that help or hinder progress. Because the roots of poverty are deep and complex, our understanding of mobility out of poverty must be too.
[3] Examples of administrative data that can be linked to surface more understanding about economic mobility include but are not limited to vital records, school attendance and achievement, housing assistance and homeless services, income supplements, and employment and earnings data.
[4] Learn more about King County’s integrated data hub here, and read more about how the hub is being leveraged to respond to the COVID-19 pandemic here.
