Collaboration by Mandate: How Shared Data Infrastructure Shapes Coordination and Control in U.S. Homelessness Services
Lingwei Cheng, Saerim Kim, Andrew Sullivan · 2026
Read Collaboration by Mandate: How Shared Data Infrastructure Shapes Coordination and Control in U.S. Homelessness Services by Lingwei Cheng, Saerim Kim, Andrew Sullivan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-07354.txt.
Introduction
Cheng, Kim and Sullivan study how mandated data sharing reshapes teamwork and control in United States homelessness services. With client consent, staff enter data into a federal system and read cross-site histories to team up on care. Yet the same standards and governance rules decide who may collect, interpret and act on data, putting decision power in fewer hands. Interviews show shared standards aid joint learning, but unequal resources push some people toward compliance roles.
Catalog record
| CREATOR | Lingwei Cheng, Saerim Kim, Andrew Sullivan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 26619 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.07354
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Collaboration by Mandate: How Shared Data Infrastructure Shapes Coordination and Control in U.S. Homelessness Services Thanks: This paper was accepted to the Extended Abstracts of the 2026 CHI Conference on Human Factors in Computing Systems (CHI EA ’26). CCS: Human-centered computing Computer supported cooperative work CCS: Human-centered computing Empirical studies in collaborative and social computing Lingwei Cheng, Saerim Kim, Andrew Sullivan Abstract . When governments mandate collaboration, shared data systems can serve both as tools for coordination and instruments of control. This study examines U.S. homelessness service networks, where Continuums of Care (CoCs) coordinate service providers through the federally mandated Homeless Management Information System (HMIS). With client consent, providers enter data into HMIS and access cross-provider service histories to support coordinated care. At the same time, HMIS embeds standards and governance rules that shape who can collect, access, interpret, and act on data, and thus who holds decision authority. Using qualitative interviews with six experts, we show that standardization can facilitate collaboration and shared learning. However, unequal resources, analytic capacity, and authority limit equitable participation and often shift some participants toward compliance-focused roles. We contribute to public-interest design research on civic data infrastructures by illustrating how mandated data sharing can simultaneously enable coordination and accountability while reproducing power asymmetries in data interpretation and decision-making. 1. Introduction Mandated collaboration — where an external authority compels organizations to coordinate through laws, regulations, and funding requirements — is common in fragmented social policy domains such as homelessness, child welfare, and disaster response ( 1 ; 15 ; 29 ) . In these domains, service delivery relies heavily on local nonprofit providers, creating incentives for government to require coordination. In the U.S., these mandates often operate through data, as federal and state governments embed data infrastructures into the core of mandated collaboration and increasingly tie funding and performance metrics to data reporting ( 20 ; 24 ) . In such settings, data become the currency of coordination, accountability, and control: data-driven collaboration can shape who receives services, how funding is allocated, how performance is assessed, and how collaborators are held accountable. In social service provision, the design of data systems can either empower nonprofit service providers as strategic collaborators or create a cycle of dis-empowerment through erosion of autonomy, data drift, and data fragmentation ( 4 ) . Policymakers, in turn, risk assuming that data mandates inherently foster collaboration when, in fact, they may reinforce power asymmetries or embed inequities in service delivery. Prior work on data and algorithms in homelessness has examined how frontline workers collect data and make decisions ( 16 ; 18 ) , and the implications for value tradeoff ( 27 ) , data quality ( 22 ) , predictive validity, fairness, and perceptions of algorithmic tools ( 21 ; 9 ) . We know less about how shared data infrastructures structure data work and decision authority across the service network. We address this gap through a case study of the U.S. homelessness system, where the federally mandated Continuums of Care (CoCs) service networks coordinate providers using a mandated shared data infrastructure: the Homeless Management Information System (HMIS). HMIS standardizes client-level data collection and reporting across service providers, enabling shared access to clients’ service histories and supporting system-wide coordination and accountability. Yet the same standards, access rules, and interpretive practices can centralize authority and exacerbate inequities between government and nonprofit providers ( 11 ; 13 ) . We ask: (1) How does a mandated shared data infrastructure structure cross-organizational data practices across the data lifecycle (collection, storage, analysis, and use)? (2) How does this infrastructure shape participation and influence? Specifically, do shared data practices support distributed, collaborative decision-making or concentrate control? Drawing on six interviews with stakeholders who manage and use HMIS in the Central Florida region, we show that effective data-driven collaboration depends not only on shared infrastructure but also on how access, analytical capacity, and interpretive authority are distributed across participants. Standardization supports interoperability and accountability, but it can limit local customization and place greater interpretive responsibility on the lead agency and more resourced organizations. Collaboration is further shaped by trust and alignment of incentives across partners. We contribute to CHI research on civic technologies and public interest data infrastructures in three ways. First, we provide an in depth qualitative study of a civic data infrastructure, HMIS, showing how it shapes collaboration, accountability, and power across government and nonprofit actors. Second, we examine not only how data is collected through shared infrastructure, but also how it is collectively analyzed and transformed into a community narrative, and how frictions such as shifts toward compliance focused roles can concentrate authority. Third, we situate homelessness related HCI research within its broader policy and organizational context, highlighting how data governance and data infrastructures jointly shape data production, interpretation, and fairness. 2. Related Work Mandated collaboration occurs when a third party requires organizations to cooperate and enforces that cooperation ( 29 ) . Mandates facilitate coordination by reducing transaction costs, excessive competition, or low inter-organizational trust. At the same time, they reshape accountability when participants are now responsible not only for their own performance but also for collective outcomes, creating potential conflicts over workload and authority. Crucially, mandated collaboration often operates through data: reporting standards, access rules, and accountability routines define performance and compliance requirements. We use data-driven collaboration to refer to cross-organizational coordination achieved through shared data practices and management. Data infrastructures shape the balance between collaborative problem-solving and hierarchical control in inter-organizational governance. In distributed systems, they enable negotiation and collective problem-solving, while centralized designs often concentrate interpretive authority and decision-making ( 2 ; 10 ) . Prior work on data use in nonprofits shows that when external stakeholders, especially funders, define performance metrics, tracking, and reporting, they can undermine nonprofit autonomy, leading to data drift, fragmentation, and disempowerment ( 4 ) . Similarly, 17 show that power in government AI adoption operates across multiple levels — within agencies, across levels of government, and downward toward clients – revealing a disconnect between frontline concerns and leadership’s valuation of AI, as well as federal guidance that overlooks local resource constraints and weakly aligns private-sector partners with public goals. Even when there is broad recognition of the value of data tools, perceived local misalignment drives demands for customization ( 28 ) . When collaboration is weak, mandates may drain resources and intensify conflict ( 6 ; 7 ) . In short, the governance embedded in data infrastructures can determine whether mandates enable collective problem-solving or instead intensify resource drain, misalignment, and conflict. The effectiveness of shared data infrastructure is shaped by resources, capacity, and organizational context. Service providers must balance direct service provision with data work ( 23 ; 20 ; 11 ) . Incomplete reporting, limited software functionality, cost barriers, and uneven technical expertise are major hurdles to effectively leverage shared data ( 25 ) . Building the social relationships and organizational capacity needed for collaborative data practice is also resource intensive ( 8 ) . When resources are scarce, participants may be unable to sustain meaningful data practices, leading to uneven participation and influence across the network ( 4 ) . Lastly, funders often tie participation and reporting to financial incentives