Pay-for-performance (P4P) is appealing because it looks simple: define metrics, pay bonuses, and expect improvement. In HCBS, poorly designed P4P often does the opposite—driving under-service, discouraging reporting, and rewarding providers who are best at documentation rather than delivery. The solution is not to abandon incentives, but to design them around real workflows, rights protections, and verifiable evidence. This article sets out a practical P4P design playbook for community services. For related foundations, see Value-Based Payment & Outcomes-Led Design and Using Data for Commissioning & Oversight.
Why “simple” pay-for-performance becomes risky in HCBS
HCBS outcomes are strongly shaped by context: housing, caregiver availability, medical fragility, and system interfaces. When P4P metrics are chosen without influenceability and verification, providers shift behavior to optimize scores rather than improve support. In the worst cases, people experience reduced contact, increased restrictive practices, and suppressed reporting because the system has paid for the wrong signals.
A defensible P4P model begins with a rights-and-safety stance: incentives must never make it financially rational to reduce service intensity or avoid complex people.
Two oversight expectations that P4P designs must meet
Expectation 1: Commissioners must show controls against under-service and inappropriate restriction
Oversight bodies increasingly expect commissioners to demonstrate that incentives do not pressure providers into reducing authorized support, discouraging escalation, or increasing restrictive practices.
Expectation 2: Performance must be verifiable and tied to real service records
Where P4P money is paid, commissioners must be able to show that performance was not achieved through coding drift or selective reporting. Verification routines are a governance requirement, not a “nice to have.”
Operational example 1: Incentivizing prevention work while protecting service intensity
What happens in day-to-day delivery: The commissioner pays incentives for verified prevention behaviors: timely post-incident debriefs, updated support plans within a defined window, and completion of high-risk reviews. At the same time, the contract requires that delivered support hours remain within an expected variance of authorized hours unless changes are documented and approved. Providers run weekly risk huddles, record actions in a structured template, and submit a monthly prevention evidence file with sampled cases.
Why the practice exists (failure mode it addresses): Incentives based on “lower utilization” can push providers to reduce support. This practice exists to pay for the governance work that prevents escalation without rewarding under-service.
What goes wrong if it is absent: Providers learn that the easiest way to improve outcomes is to reduce contact or delay escalation. People then deteriorate until crises occur, undermining safety and increasing system costs.
What observable outcome it produces: Prevention incentives produce more consistent risk management. Evidence includes timely care plan updates, documented debriefs, stable delivered-hour patterns, and reduced repeat incidents for high-risk members.
Operational example 2: Anti-gaming verification using record sampling and exception controls
What happens in day-to-day delivery: Commissioners implement a monthly verification routine: a small random sample of incentivized cases is reviewed against source records (case notes, incident logs, EVV/scheduling, care plan timestamps). Providers must maintain an audit folder for sampled cases. Exceptions (e.g., hospitalization, relocation) require documented approval using a standardized exception form with time limits and evidence requirements.
Why the practice exists (failure mode it addresses): If incentives can be won by redefining exclusions or improving documentation alone, they will be. This practice exists to prevent “paper performance” and to maintain trust that incentives reflect real delivery.
What goes wrong if it is absent: Providers improve scores through selective reporting or exception inflation. Commissioners later discover issues through complaints, adverse events, or financial audits, leading to clawbacks and damaged relationships.
What observable outcome it produces: Verification reduces disputes and improves data quality. Evidence includes decreasing discrepancy rates, stable exception volumes with documented approvals, and clearer links between incentive payments and real practice changes.
Operational example 3: Rights-based safeguards that stop incentives driving restrictive practice drift
What happens in day-to-day delivery: The contract includes a rights safeguard: any provider with rising restrictive practice indicators (e.g., increased PRN use, seclusion-related events where relevant, or restrictive interventions documented in support plans) triggers an automatic review, and incentive payment is paused pending assurance. Providers must evidence least-restrictive practice governance: behavior support plan review cadence, de-escalation training completion, and supervisory oversight of high-risk interventions.
Why the practice exists (failure mode it addresses): Incentives tied to “stability” can unintentionally reward containment rather than support. This practice exists to prevent restrictive practices becoming the pathway to metric improvement.
What goes wrong if it is absent: Providers may keep “incidents” down by using restrictive interventions more often, discouraging reporting, or limiting community exposure—creating rights violations that surface later through serious incidents or oversight findings.
What observable outcome it produces: Rights safeguards maintain ethical practice while still allowing performance incentives. Evidence includes stable or reduced restrictive practice indicators, documented review minutes, and improved member experience signals without suppressed incident reporting.
Closing: pay for verifiable practice improvements, not just numbers
P4P can work in HCBS when incentives target influenceable practices, under-service is actively monitored, and verification is built into the operating model. The goal is not to “score well,” but to fund the behaviors that make safe, rights-respecting outcomes achievable.