Managing Strain Rather Than Beds: Dynamic Capacity and Adaptive Response In Foster Care Systems

Background and Purpose: Child welfare systems must maintain sufficient foster home capacity to place children safely while avoiding excess, poorly targeted supply. This challenge is increasingly framed within expectations for system performance, timely placement, and demonstrable capacity. We reconceptualize foster care capacity not as a static inventory problem (i.e., beds), but as a dynamic strain-management problem shaped by the interaction of child demand, foster home supply, and the system’s ability to adapt over time. Rather than asking how many homes are needed, the central question becomes whether the system can absorb demand shocks without entering periods of system strain.

Methods: Using weekly administrative data from two jurisdictions with parallel reporting structures, we analyzed foster home supply, openings and closings, child population, and child admissions and discharges over five years. We constructed a capacity ratio (homes per child) and examined the ratio across annual, quarterly, monthly, and weekly time scales. In addition to descriptive analysis, we assessed admissions and discharge patterns, evaluated relationships between demand and supply response, and examined the extent to which fluctuations in child admission were matched by changes in foster home availability. These analyses were used to characterize the system’s exposure to forward-looking strain, defined as the risk that capacity will become insufficient over a short planning horizon.

Results: The two jurisdictions exhibited similar seasonal patterns but substantially different operating margins. In the first jurisdiction, the capacity ratio remained consistently above parity, indicating a buffered system with available slack. In the second, the ratio hovered near the threshold of constraint, reflecting a tighter system more vulnerable to demand fluctuations. Across both jurisdictions, increases in child admissions were only weakly matched by increases in foster home supply. Lagged analyses indicated modest and diffuse supply responses, with no strong evidence of rapid or sustained adjustment following demand shocks. Additional analyses showed that, in one jurisdiction, admissions behaved approximately as a random process, while in the other they exhibited a persistent pattern, implying differing degrees of predictability. In both settings, demand increases tended to tighten capacity more quickly than supply could respond, resulting in recurring periods of elevated strain.

Conclusions and Implications: These findings shift the policy focus from static bed counts to dynamic alignment between demand and supply over time. For system leaders, effective capacity management requires monitoring not only supply levels, but also the probability of entering constrained states, the duration and recurrence of strain, and the lag between demand shocks and supply response. Strain should be understood as a forward-looking risk condition, not simply a contemporaneous shortage. For researchers, foster care capacity should be treated as an endogenous system property that both shapes and is shaped by service processes, rather than as a fixed backdrop to placement outcomes. More broadly, the results suggest that improving child welfare performance depends on strengthening the system’s adaptive capacity—its ability to anticipate, absorb, and respond to fluctuations in demand—rather than simply increasing the number of available homes.