Stakeholder Incentive Analysis: Designing Programs for Misaligned Actors

Module 7: Policy, Incentives, and Public Systems Behavior Depth: Application | Target: ~1,500 words

Thesis: Every stakeholder in a public program — funder, grantee, provider, patient, regulator — has incentives that may or may not align with program goals, and stakeholder incentive analysis is prerequisite to program design.

The Design Problem

Program designers routinely treat stakeholder analysis as a political exercise — mapping who supports the initiative, who might resist, who needs to be consulted. This is necessary but insufficient. The harder problem is incentive analysis: understanding what each stakeholder is optimizing for, determining where those optimization targets diverge from program goals, and designing the program so that the divergences do not destroy its effectiveness.

The distinction matters because political support and incentive alignment are different things. A stakeholder can enthusiastically endorse a program while holding incentives that systematically undermine it. A grantee organization may genuinely want to expand opioid treatment capacity — and simultaneously optimize for grant revenue stability in ways that prioritize easily documented outputs over difficult-to-measure access improvements. The endorsement is sincere. The incentive misalignment is structural. Program design that addresses only the political dimension and ignores the incentive dimension will produce programs that enjoy broad support and deliver modest results.

Stakeholder incentive analysis, as formalized by Freeman (1984) in stakeholder theory and extended through principal-agent economics (Jensen and Meckling, 1976; Eisenhardt, 1989), treats every actor in a system as a rational agent with distinct objectives, constraints, and information. The analytical question is not “does this stakeholder support the program?” but “what will this stakeholder do when the program’s goals conflict with their own incentives — and where will those conflicts concentrate?”

Five Stakeholder Categories in Healthcare Grants

Public healthcare programs operate through a chain of actors, each with distinct incentive structures that interact in predictable ways.

Funders (federal agencies — HRSA, SAMHSA, CMS — and state pass-through entities) optimize for portfolio-level impact and political accountability. A federal program officer managing 40 grants is evaluated on program-wide outcomes: total treatment capacity created, geographic coverage, aggregate performance metrics. The funder wants measurable, reportable impact across the portfolio. Individual grantee complexity is noise. This creates an incentive for standardized metrics and uniform reporting — which may or may not capture what matters at any individual site. The funder also faces appropriations risk: demonstrating program-wide success is necessary to justify continued Congressional funding. The incentive to report portfolio success can create tolerance for individual grantee metric gaming, because challenging a grantee’s reported numbers threatens the portfolio narrative.

Grantee organizations (health systems, FQHCs, behavioral health agencies, tribal health organizations) optimize for financial sustainability and compliance. The grant is revenue. The grantee needs the revenue to continue, which means meeting milestone requirements and maintaining compliance standing for future awards. This creates dual optimization: deliver enough program results to justify continuation, and maintain the compliance documentation that prevents audit findings. When these objectives conflict — when genuine program adaptation requires deviating from approved milestones, or when honest outcome reporting would reveal underperformance — the grantee faces a structural tension between program fidelity and organizational survival. Organizational survival usually wins. The grantee reports what the funder needs to see, which may diverge from what is actually happening on the ground.

Frontline providers (clinicians, care coordinators, outreach workers) optimize for manageable workload and professional effectiveness within the constraints they face. A provider assigned to a grant-funded opioid treatment expansion does not optimize for “treatment capacity” in the abstract. She optimizes for a caseload she can manage competently, documentation she can complete within her shift, and patient interactions that align with her clinical training. When program goals require intake volume that exceeds what she considers clinically manageable, she will — rationally and often unconsciously — adjust. Intake screening becomes more selective. Referral criteria tighten. The provider is not sabotaging the program. She is managing the gap between what the program demands and what she can deliver at the quality level her professional identity requires.

Patients and community members optimize for access, convenience, and perceived quality of care. A patient seeking opioid treatment wants treatment she can get to, at a time she can attend, from a provider she trusts, with minimal bureaucratic friction. She does not care about the grant’s milestone structure. She does not know that the program has a capacity target. She experiences the program as the wait time for an intake appointment, the number of forms she must complete, the distance to the clinic, and whether the provider treats her with dignity. When program design creates friction — complex intake processes designed to generate documentable screening data, inconvenient hours driven by staffing constraints, eligibility requirements shaped by the grant’s target population definition — the patient experiences a barrier. If the barrier is high enough, she does not access the program. The funder sees capacity created. The patient sees a door she cannot get through.

Regulators and auditors (OIG, state oversight agencies, single audit firms) optimize for documentation completeness and regulatory adherence. An auditor evaluating a SAMHSA grant is not assessing whether patients received effective treatment. She is assessing whether the grantee can demonstrate, through documented evidence, that funds were spent in accordance with 2 CFR 200, that reported activities occurred as described, and that financial controls were maintained. The auditor’s incentive is to identify discrepancies between documentation and requirements — not to evaluate program effectiveness. This creates a system where the grantee’s compliance effort is directed toward satisfying the auditor, which may or may not correlate with program quality. A program with impeccable documentation and mediocre outcomes passes audit. A program with strong outcomes and sloppy documentation generates findings.

The Principal-Agent Chain

These five stakeholders form a principal-agent chain: funder delegates to grantee, grantee deploys providers, providers serve patients, regulators oversee the entire chain. Eisenhardt (1989) identified the core problem of agency theory as information asymmetry: the principal cannot directly observe the agent’s effort, only its outcomes — imperfectly and with delay. In a single principal-agent relationship, this creates moral hazard and adverse selection (see WF M4, 04-incentive-alignment.md). In a chain with multiple links, each link introduces its own information asymmetry, and the distortions compound.

The funder cannot observe grantee operations directly. She sees quarterly reports. The grantee cannot observe provider behavior directly. She sees EHR documentation and productivity metrics. The provider cannot observe patient experience directly. She sees appointment attendance and survey responses. At each link, the principal relies on metrics that the agent controls, creating the conditions for Goodhart’s Law to operate at every level simultaneously.

Where misalignment concentrates. In practice, the most damaging misalignments in healthcare grant programs cluster at two links: grantee-to-provider and provider-to-patient.

The grantee-provider link fails because the grantee’s optimization target (milestone compliance, grant continuation) translates into provider directives (intake volume targets, documentation requirements) that may conflict with clinical judgment. The provider responds by adjusting behavior within her discretion — tightening intake criteria, extending appointment intervals, simplifying screening protocols — in ways that are invisible to the grantee but reduce program access or quality. The grantee sees milestone metrics. She does not see the patients who were screened out or the clinical corners that were cut.

The provider-patient link fails because the provider’s caseload management decisions directly determine patient access, and the patient has no mechanism to signal unmet need within the program’s reporting structure. A patient who calls for an intake appointment and is told the wait is six weeks does not appear in the program’s data as a failure. She appears nowhere. The program reports treatment capacity. It does not report the population that tried to access that capacity and could not.

Diagnosing Misalignment Before It Manifests

The SAMHSA opioid treatment expansion example makes the diagnostic pattern concrete. The funder awards a three-year grant to a rural behavioral health agency to expand medication-assisted treatment (MAT) capacity. The grant specifies milestones: hire two additional MAT providers by month six, establish 60 new treatment slots by month twelve, achieve 80% slot utilization by month eighteen.

The misalignments are predictable from the incentive structure:

The grantee reports capacity created (slots established) because that is the milestone. Slot utilization — whether those slots are actually filled with patients receiving treatment — is a secondary metric that requires more effort to document and may reveal uncomfortable truths about access barriers. The grantee optimizes for what the funder measures first.

The providers reduce intake complexity to manage caseload. The grant envisions comprehensive intake assessments — trauma screening, co-occurring disorder evaluation, social determinants documentation — because the funder wants evidence-based treatment. The providers, managing a growing caseload with no increase in administrative support, abbreviate intake. The screening data that the program was designed to generate becomes thin or formulaic. The clinical value of comprehensive intake erodes.

The patients experience the program as the wait time for the first appointment, the distance to the clinic, and the complexity of the paperwork. Rural patients with opioid use disorder face transportation barriers, employment constraints, and stigma. Every additional friction point in the intake process is a dropout opportunity. The program’s design — shaped by the funder’s evidence requirements and the grantee’s documentation needs — creates friction that the target population is least equipped to navigate.

The funder sees milestone compliance: providers hired, slots created, utilization within an acceptable range. The actual access improvement — the number of previously untreated patients now in sustained recovery — is modest relative to the investment. The program is compliant. The program is underperforming. And nobody in the chain has the incentive to surface the gap.

Diagnostic questions for operators. Before program launch, ask at each link in the chain:

What is each stakeholder optimizing for, independent of the program’s stated goals?
Where does the metric the principal sees diverge from the behavior the agent controls?
What will the agent do when program goals conflict with their workload, revenue, or compliance incentives?
What patient experience data exists that is not captured by the program’s reporting structure?
Which milestone metrics can be satisfied without producing the intended outcome?

These questions are uncomfortable. They surface the structural tensions that polite stakeholder engagement avoids. But they are prerequisite to program design that accounts for misalignment rather than discovering it in year-three evaluation findings.

Integration Points

Workforce Module 4: Incentive Alignment (WF M4, 04-incentive-alignment.md). WF M4 analyzes incentive misalignment within organizations — how compensation structures, productivity metrics, and non-financial motivators shape clinician behavior. The stakeholder incentive analysis in this module extends that framework across organizational boundaries. The provider in a grant-funded program faces incentives from two principals simultaneously: the employing organization (productivity, documentation, compliance) and the grant program (treatment volume, outcome metrics, evidence generation). When these principal demands conflict — when the organization wants efficient throughput and the grant wants comprehensive screening — the provider is caught between optimization targets. The moral hazard and adverse selection dynamics described in WF M4 operate at each link in the principal-agent chain, compounding across the multi-stakeholder system that grant programs create.

Human Factors Module 8: Incentive Gaming (HF M8, 08-incentive-gaming.md). Gaming is the behavioral mechanism through which stakeholder misalignment manifests in program metrics. When a grantee reports slot capacity rather than slot utilization, she is not committing fraud — she is optimizing the metric the funder measures, exactly as Goodhart’s Law predicts. When a provider tightens intake criteria to manage caseload, she is engaging in the cherry-picking behavior that HF M8 taxonomizes. The stakeholder incentive analysis in this module identifies where gaming will concentrate; HF M8 identifies how it will manifest and what metric design resists it. Program designers who conduct incentive analysis before launch can predict the gaming pathways and design metrics that close them — rather than discovering gaming in year-two evaluation data and attributing it to individual bad actors rather than structural incentive misalignment.

Product Owner Lens

What is the funding/compliance/execution problem? Each stakeholder in the grant chain optimizes for different objectives. The program’s reporting structure captures the funder’s metrics but not the misalignment between those metrics and on-the-ground behavior. The gap between reported performance and actual impact is structurally invisible.

What mechanism explains the operational bottleneck? Principal-agent theory across a multi-link chain (Jensen and Meckling, 1976; Eisenhardt, 1989). Each link introduces information asymmetry; each asymmetry creates space for the agent’s optimization to diverge from the principal’s intent. Misalignment compounds across links, with the greatest distortion at the grantee-provider and provider-patient boundaries.

What controls or workflows improve it? Pre-launch stakeholder incentive mapping as a required program design step. Dual-metric reporting that pairs output metrics (capacity created) with access metrics (capacity utilized, patient-reported access barriers). Provider-level feedback loops that surface caseload management decisions to program managers before they calcify into access restrictions.

What should software surface? A stakeholder incentive map that visualizes each actor’s optimization target alongside the program’s intended outcome — making divergences explicit. Utilization-to-capacity ratios tracked at the provider level, not just the program level, to detect where access is being rationed. Patient journey tracking from initial contact through treatment engagement, surfacing the dropout points that milestone reports do not capture. Alert when output metrics (slots, hires, trainings) advance while outcome metrics (utilization, retention, patient-reported access) lag — the divergence pattern that signals misalignment is operating.

What metric reveals risk earliest? The ratio of capacity reported to capacity utilized, disaggregated by provider. When aggregate utilization looks acceptable but provider-level data shows high variance — some providers at 95% utilization and others at 40% — the variance signals that intake gatekeeping is uneven and likely driven by individual provider workload management rather than patient demand. This variance is the earliest operational signal that the grantee-provider misalignment is active.

Warning Signs

These indicators suggest stakeholder misalignment is degrading program performance:

Output milestones (capacity created, staff hired, trainings completed) are met on schedule while outcome metrics (utilization, retention, patient access) lag or are unreported
Providers report caseload pressure while program-level data shows slots available — a sign that intake gatekeeping is filtering demand the program was designed to serve
Quarterly reports describe activities and deliverables but do not address whether the target population is being reached
Patient complaints or community feedback describe access barriers that do not appear in program reporting
Grantee requests no-cost extensions or budget modifications that shift resources from direct service to documentation and compliance infrastructure
Program evaluation finds “implementation challenges” without tracing those challenges to specific stakeholder incentive conflicts
Auditors find no compliance issues in a program with mediocre outcomes — clean documentation masking weak impact