System Dynamics of Housing First and Medications for Opioid Use Disorder: Stock-and-Flow Quantifications, Buffering Efficacy, and Archetypal Policy Feedback

1. Methodological and System Dynamics Framework

Public health and social welfare policies frequently confront complex human crises characterized by high morbidity, acute mortality, and institutional churn. Within modern social policy, low-barrier harm reduction models have emerged as primary paradigms for stabilizing vulnerable populations. In the housing sector, this is operationalized through Housing First (HF)—predominantly executed via Permanent Supportive Housing (PSH) and Rapid Re-Housing (RRH)¹. In addiction medicine, it is operationalized through Medications for Opioid Use Disorder (MOUD), which utilizes opioid full agonists (methadone), partial agonists (buprenorphine), and antagonists (extended-release naltrexone)².

To evaluate the mechanical operation, clinical efficacy, and macro-level consequences of these strategies, they must be situated within the mathematical and conceptual architecture of system dynamics. In classical system dynamics, a system’s state is captured through accumulations known as stocks, which represent the physical, biological, or human entities that provide the system with memory, inertia, and delays. Transitions into and out of these accumulations are governed by flows—rates of change measured over discrete time increments:

Within this analytical framework, populations experiencing unsheltered homelessness or active, unregulated opioid use disorder (OUD) constitute distressed upstream stocks. Housing First programs and MOUD regimens operate as specialized intermediate stocks intended to intercept these populations, mitigate acute physiological and social hazards, and alter downstream public costs⁴.

A critical evaluative question for policy design is whether an intervention acts as a fundamental curative transformation or primarily as an operational buffer. A fundamental curative transformation alters the internal state variables of the individual—such as permanently resolving psychiatric pathology, reversing neurochemical dependence, or building durable socio-economic independence—thereby ensuring that the individual remains stable even if external programmatic support ceases. Conversely, an operational buffer acts as an external stabilizing force that shields the individual from acute mortality and environmental trauma exclusively while they remain physically or pharmacologically retained within that stock. Once the buffer is removed, the individual’s baseline biological or environmental vulnerability reasserts itself.

When public systems rely disproportionately on operational buffers to suppress immediate crises while neglecting the causal drivers of those crises, the systemic feedback frequently activates the “Shifting the Burden” system archetype⁶. In this archetype, a symptomatic balancing loop rapidly mitigates visible distress, eroding the political and financial urgency required to implement fundamental balancing solutions that carry extended delays⁶. Over time, reinforcing side-effect loops undermine fundamental system capacity, trapping public infrastructure in an escalating reliance on symptomatic containment⁶.

2. Stock and Flow Architecture of Housing First and Permanent Supportive Housing

2.1 Characterization of National Stocks and Capacity Reserves

The United States homeless response system functions as a network of interconnected accumulations distributed across street environments, temporary emergency facilities, and permanent subsidized programs¹. The distressed human population is partitioned into two primary upstream stocks: the unsheltered homeless stock, consisting of individuals residing in places not meant for human habitation (such as encampments, vehicles, and abandoned buildings), and the sheltered homeless stock, comprising individuals residing in emergency shelters or transitional facilities¹. Based on Point-in-Time (PIT) counts compiled in the Department of Housing and Urban Development (HUD) Annual Homeless Assessment Report (AHAR) to Congress, the national single-night cross-section of unhoused individuals exceeds 650,000, with more than 140,000 meeting the federal definition of chronic homelessness—characterized by a long-term disabling condition combined with continuous homelessness for at least twelve months or multiple episodes totaling a year over a three-year period¹.

Parallel to these population stocks are the institutional housing capacity stocks funded through HUD Continuums of Care (CoCs) and the Department of Veterans Affairs (VA)⁹:

The Permanent Supportive Housing (PSH) Stock represents long-term subsidized units linked with voluntary, wraparound supportive services, designed for individuals with chronic physical, psychiatric, or substance use disabilities¹. Nationally, PSH capacity expanded to over 380,000 dedicated beds, with roughly 204,000 adult and veteran households residing within these placements at any given time¹.
The Rapid Re-Housing (RRH) Stock represents time-limited rental assistance paired with basic housing stabilization services, holding approximately 90,700 adult households in use concurrently¹.
The Transitional Housing (TH) Stock represents programmatic, facility-based congregate housing requiring service compliance. As federal funding prioritized Housing First approaches over the past decade, TH was systematically contracted, resulting in the elimination of more than 120,000 transitional beds nationally¹³.

2.2 Transition Flows: Placement Velocities, Retention Half-Lives, and Churn

The movement of individuals through these housing stocks is characterized by severe placement bottlenecks, prolonged retention, and constrained outflows:

The primary placement inflow into permanent housing stocks is severely restricted by overall bed availability. National Homeless Management Information System (HMIS) data shows that homeless response systems have the aggregate capacity to place only 16.0% of the households served in emergency shelters into permanent housing within a given year¹. For non-veteran single adults experiencing chronic homelessness, the annualized placement rate into PSH satisfies only 13.4% of total demand¹. By contrast, chronically homeless veterans access PSH at a 47.9% rate, and chronically homeless families at a 45.5% rate, reflecting dedicated federal resource allocations through programs such as HUD-VASH¹.

Once an individual transitions into the PSH stock, retention is exceptionally durable. Across national programs, 84% to 86% of participants maintain their housing for at least twelve months¹². In multi-site experimental trials, individuals placed into Housing First spend between 80% and 85% of their days stably housed over multi-year observation windows, compared to 60% or less among individuals navigating treatment-first or standard municipal systems¹⁵.

Because PSH is structured as a permanent subsidy without mandatory exit criteria, its outflow rate is low. Adult and veteran households exit PSH at an annual rate of 12.8%, while family households exit at 13.2%¹. At these turnover velocities, it requires approximately 7.8 years for the existing adult PSH stock to completely turn over and reopen its capacity for new entrants¹. RRH, which provides temporary rental assistance, functions with a much higher outflow velocity, exhibiting an annual exit rate of 64.9% for adult households and a total capacity replenishment period of 1.5 years¹.

The programmatic outflows from PSH diverge into four discrete terminal destinations:

Transition into unassisted, private-market permanent housing, which occurs in only a small fraction of chronic cases due to ongoing poverty and disability¹.
Institutional placement into skilled nursing facilities or medical long-term care⁶.
Mortality within the housing unit. HUD administrative data reveals that 19.5% of all exits from PSH among single adults living alone are due to tenant death, underscoring the severe medical vulnerability of this chronic cohort¹⁷.
Returns to homelessness. Longitudinal performance tracking through HUD System Performance Measures (Measure 2) shows that between 10% and 16% of individuals exiting permanent housing programs return to emergency shelters or unsheltered street environments within 24 months, with point-in-time assessments in major urban centers revealing that approximately 10% of street-based populations had previously received PSH¹³.

2.3 Empirical Outcomes Pre- and Post-Entry

The transition of chronically unhoused individuals into Housing First stocks substantially alters their interaction with acute public infrastructure, as detailed in the following data synthesis:

3. Stock and Flow Architecture of Medications for Opioid Use Disorder (MOUD)

3.1 Characterization of National Epidemiological Stocks

The epidemiological landscape of opioid use disorder in the United States is structured by a large untreated population and constrained clinical harm reduction stocks². Epidemiological data from the National Survey on Drug Use and Health (NSDUH) indicates that over 6.1 million individuals aged 12 and older meet diagnostic criteria for past-year OUD, with state-level per-capita prevalence spanning 0.75% to nearly 3.0% of the adult population².

This untreated population interacts with three primary medication maintenance stocks, each constrained by distinct regulatory and operational structures²:

The Methadone Stock functions as a full mu-opioid receptor agonist administered exclusively through specialized, highly regulated Opioid Treatment Programs (OTPs) subject to strict federal dosing and attendance mandates².
The Buprenorphine Stock operates as a partial mu-opioid agonist with high receptor affinity and slow dissociation kinetics, delivered primarily through office-based general medical practices, community clinics, and specialized outpatient settings²⁵. The regulatory capacity for buprenorphine shifted in January 2023 following the passage of Section 1262 of the Consolidated Appropriations Act (the MAT Act), which eliminated the federal DATA-2000 “X-waiver” requirement, thereby expanding the potential prescriber pool to over 53,600 clinicians²⁵.
The Extended-Release Naltrexone Stock consists of a monthly intramuscular depot of a full mu-opioid receptor antagonist, requiring full physiological detoxification and complete opioid abstinence prior to induction².

Despite the pharmacological efficacy of these medications, treatment penetration remains low. Across the national OUD stock, fewer than 20% to 25% of individuals receive any MOUD in a given year²⁵. Following a sentinel nonfatal opioid overdose treated in an emergency setting, only 11% of patients enroll in methadone maintenance, 17% initiate buprenorphine, and 6% receive naltrexone within the subsequent twelve months, leaving more than two-thirds of high-risk overdose survivors entirely disconnected from pharmacotherapy³.

3.2 Transition Flows: Initiation Velocities, Retention Curves, and Attrition

The flow dynamics of MOUD are defined by modest initiation velocities and steep attrition curves:

The rate of transition from the untreated OUD stock into pharmacotherapy is constrained by clinical hurdles and provider capacity. Following acute inpatient medically managed withdrawal (MMOW), the 28-day initiation rate is just 13.4% for buprenorphine and 4.5% for extended-release naltrexone⁵. While the statutory removal of the X-waiver induced an immediate addition of 1,900 new prescribers, real-world patient initiation rose by only 5,245 patients in the first month nationwide, demonstrating that clinical comfort, administrative overhead, and social stigma continue to restrict entry flows²⁵.

Once patients initiate medication, retention half-lives vary systematically across pharmacological classes. In longitudinal post-overdose cohorts, the median duration of continuous treatment retention is 5 months for methadone (interquartile range [IQR]: 2 to 9 months), 4 months for buprenorphine (IQR: 2 to 8 months), and 1 month for extended-release naltrexone (IQR: 1 to 1 month)³.

The CDC longitudinal MOUD cohort demonstrates steady attrition across all outpatient modalities: retention declines from 100% at baseline to 72% at 3 months, 68% at 6 months, and 52% to 53% at 12 to 18 months²³. Across clinical trials, methadone maintenance exhibits higher retention rates than buprenorphine, while naltrexone suffers from severe induction dropouts and high discontinuation rates⁵.

3.3 Clinical and Epidemiological Outcomes: On-Treatment Versus Out-of-Treatment

When patients transition into active MOUD maintenance stocks, their risk profiles improve substantially across multiple clinical and public health dimensions:

Clinical / Mortality Dimension	Out of Treatment (Active Unregulated OUD)	Active Treatment: Methadone Maintenance	Active Treatment: Buprenorphine Maintenance	Pooled Treatment Effect Size (RR / AHR)	Primary Source
All-Cause Mortality Rate (per 1,000 Person-Years)	20.98 to 36.1 deaths per 1,000 person-years	9.09 to 11.3 deaths per 1,000 person-years	4.3 to 9.5 deaths per 1,000 person-years	Combined OAT RR: 0.47 (0.42–0.53); Methadone: 0.47; Buprenorphine: 0.34	Sordo et al.³⁰; Santo et al.³³
Overdose Mortality Rate (per 1,000 Person-Years)	12.7 per 1,000 PY (Methadone cohorts); 4.6 per 1,000 PY	2.6 deaths per 1,000 person-years	1.4 deaths per 1,000 person-years	Out-to-In Rate Ratio: Methadone: 4.80 (2.90–7.96); Buprenorphine: 2.20	Sordo et al. Meta-Analysis³⁰
Post-Overdose All-Cause Mortality (12-Month Cohort)	Reference Category: 4.7 deaths per 100 person-years	Adjusted Hazard Ratio (AHR): 0.47 (95% CI: 0.32–0.71)	Adjusted Hazard Ratio (AHR): 0.63 (95% CI: 0.46–0.87)	Methadone: 53% reduction; Buprenorphine: 37% reduction; Naltrexone: AHR 1.44 (not significant)	Larochelle et al.³
Post-Overdose Opioid Mortality (12-Month Cohort)	Reference Category: 2.1 deaths per 100 person-years	Adjusted Hazard Ratio (AHR): 0.41 (95% CI: 0.24–0.70)	Adjusted Hazard Ratio (AHR): 0.62 (95% CI: 0.41–0.92)	Methadone: 59% reduction; Buprenorphine: 38% reduction	Larochelle et al.³
Nonfatal Overdose Recurrence (Insured OUD Cohort)	Reference Category: Detoxification or behavioral care alone	Adjusted Hazard Ratio (AHR): 0.24 at 3 months	Adjusted Hazard Ratio (AHR): 0.41 at 12 months	76% reduction at 3 months; 59% reduction at 12 months (MOUD was the only effective pathway)	Wakeman et al.²⁵
Specific Non-Overdose Causes of Death (RR During Tx)	High baseline mortality across medical conditions	Cardiovascular mortality RR: 0.69 (0.60–0.79)	Suicide mortality RR: 0.48 (0.37–0.61); Cancer RR: 0.72	Statistically significant reduction across multiple medical causes	Santo et al.³³

4. Empirical Evaluation of the Buffering Hypothesis

A key question in evaluating harm reduction policies is whether an intervention functions as an internal curative transformation or as an operational buffer. Curative interventions permanently alter individual internal capacities, leaving the person resilient even if programmatic supports end. An operational buffer, by contrast, shields the individual from environmental, legal, or biochemical hazards strictly while they are retained within the intervention stock. The empirical evidence indicates that both Housing First and MOUD operate primarily as operational buffers rather than curative interventions.

4.1 Housing First: Spatial Stabilization Versus Psychiatric and Substance Use Trajectories

The core theoretical hypothesis of the Housing First movement was that immediate, unconditional residential stability would relieve survival stress, leading to secondary reductions in psychiatric symptom severity and problematic substance use¹⁵.

Multi-site randomized controlled trials and rigorous systematic reviews have consistently disproven this secondary hypothesis. When comparing Housing First directly to Treatment-as-Usual (TAU) or Treatment-First approaches, Housing First achieves superior housing retention, but it produces virtually no differentiated clinical reduction in substance use severity or psychiatric pathology⁴.

In the Canadian At Home/Chez Soi randomized trial—the largest controlled study of Housing First to date (N > 2,000)—participants were randomized to Housing First (supported by Assertive Community Treatment or Intensive Case Management) or local municipal TAU¹⁵. Over follow-up periods extending from two to six years, Housing First participants achieved markedly higher housing stability (85% of days stably housed versus 60% in TAU)¹⁵. However, longitudinal measurements demonstrated no statistically significant differences between Housing First and TAU in substance use severity, overall alcohol consumption, or illicit drug expenditures¹⁵. Both groups showed modest, identical reductions over time, primarily reflecting regression to the mean and general contact with study personnel³⁷.

Parallel findings emerge across mental health measures. Standardized psychiatric symptom evaluations using the Positive and Negative Syndrome Scale (PANSS) for psychotic symptomatology and the Colorado Symptom Index (CSI) for broad psychiatric distress showed equivalent trajectories across both study arms⁴. The CDC Community Guide Systematic Review (Peng et al., 2020) confirmed that the relative difference in mental health symptom severity between Housing First and TAU was negligible (-2%, interquartile interval: -5% to +4%)⁴. Furthermore, illegal drug use outcomes showed an unfavorable or negligible relative difference (+11% to +62%) compared to treatment-contingent models⁴.

Longitudinal group-based trajectory analyses from the Toronto At Home/Chez Soi cohort (Kirst et al.) demonstrated that participant recovery pathways were dictated not by housing status, but by underlying comorbidities: histories of adverse childhood experiences, traumatic brain injuries, and chronic psychiatric illness⁴⁰. Receiving Housing First was predictive of perceived life quality, but persistent psychiatric morbidity and substance use disorders remained stable over time⁴⁰.

Housing First functions as a spatial and physical buffer. By moving an individual from the street into an apartment, it isolates them from the acute survival traumas of the outdoor environment—such as hypothermia, physical victimization, and crisis-driven emergency room visits⁴. However, it does not cure underlying neurochemical dependencies or psychiatric disorders; clients remain chronically mentally ill or substance-dependent, but are supported indoors⁴.

4.2 Medications for Opioid Use Disorder: Receptor Maintenance Versus Post-Cessation Mortality Dynamics

A parallel dynamic governs Medications for Opioid Use Disorder. Public discourse often treats MOUD as a rehabilitative medical cure that resolves addiction after a finite course of treatment. Epidemiological mortality data directly contradicts this premise. MOUD reduces all-cause and overdose mortality strictly during active pharmacological maintenance; once an individual transitions off the medication, the protective effect disappears, and mortality risk rises substantially³.

The pooled meta-analyses of international observational cohorts by Sordo et al. (BMJ 2017) and Santo et al. (JAMA Psychiatry 2021) document mortality patterns across the MOUD treatment lifecycle³⁰:

The absence of any legacy protective effect is evident in raw mortality differentials. Individuals who discontinue methadone maintenance die at an all-cause rate of 36.1 per 1,000 person-years, compared to 11.3 while retained in treatment (out-to-in rate ratio: 3.20)³⁰. For buprenorphine, the mortality rate is 9.5 per 1,000 person-years out of treatment versus 4.3 in treatment (rate ratio: 2.20)³⁰. The disparity is even more pronounced for overdose-specific mortality: individuals out of methadone treatment die from fatal drug poisonings at 4.80 times the rate of those actively retained (12.7 vs. 2.6 per 1,000 person-years)³⁰.

The period immediately following treatment discontinuation represents a window of heightened vulnerability:

Sordo et al. found that during the first four weeks following treatment cessation, all-cause mortality rises to approximately 32.0 deaths per 1,000 person-years across both methadone and buprenorphine cohorts³⁰.
Santo et al. established that during the four weeks post-cessation, all-cause mortality is 6.01 times higher (RR: 6.01; 95% CI: 4.32 to 8.36) than during active treatment maintenance³³.
Once this four-week window passes, mortality risk does not normalize to baseline community levels; it remains elevated at nearly double the active treatment rate (RR: 1.81; 95% CI: 1.50 to 2.18) for the remaining duration that the individual is off pharmacotherapy³³.

This post-cessation mortality spike is driven by basic neuropharmacological mechanisms. Continuous agonist or partial-agonist therapy downregulates mu-opioid receptors while maintaining biological tolerance to fatal respiratory depression²⁶. When treatment is interrupted—due to voluntary exit, administrative discharge, incarceration, or financial disruption—opioid tolerance declines rapidly²⁶. If the individual subsequently relapses into an illicit drug supply dominated by illicitly manufactured fentanyl, they lack the cross-tolerance needed to survive their prior dosage levels, precipitating fatal respiratory failure².

A secondary period of elevated risk occurs during methadone induction: all-cause mortality and overdose poisoning are approximately twice as high in the first four weeks of treatment as in subsequent maintenance phases (RR: 2.01; 95% CI: 1.55 to 5.09), reflecting the risks of drug accumulation prior to achieving pharmacokinetic steady-state³⁰. Buprenorphine, because of its partial agonist ceiling effect, displays no elevated induction mortality (RR: 0.58; 95% CI: 0.18 to 1.85)³⁰.

MOUD functions as an active pharmacological buffer. It binds receptors, normalizes the stress response, relieves cravings, and provides cross-tolerance against overdose solely while the medication is present in the bloodstream². It does not resolve underlying psychosocial trauma, economic distress, or psychiatric comorbidities². Once discontinued, the individual returns to a hazardous environment with diminished physiological protection.

5. Empirical Verification of the “Shifting the Burden” Archetype

The system dynamics archetype “Shifting the Burden” (also referred to as “Addiction to the Intervenor”) describes a system where an underlying problem generates acute symptoms⁶. The system implements an accessible, symptomatic solution that rapidly attenuates the visible distress (Balancing Loop B1). This apparent relief reduces the systemic urgency to fund and implement fundamental solutions (Balancing Loop B2), which require larger investments and involve extended delays. Crucially, the symptomatic intervention produces unintended side effects (Reinforcing Loop R1) that erode the system’s ability to ever implement the fundamental solution, creating an escalating dependency on the symptomatic fix⁶.

5.1 Housing First and the Displacement of Macro-Housing Supply

In the housing and homelessness domain, the Shifting the Burden dynamic operates across municipal and national policy environments:

The Problem Symptom consists of visible street encampments, rising Point-in-Time counts, emergency department overcrowding, and public costs tied to chronic street homelessness⁶.

The Symptomatic Solution (Loop B1) is the deployment of Housing First through dedicated Permanent Supportive Housing and Rapid Re-Housing¹. By placing high-need individuals into subsidized housing without preconditions, municipalities rapidly remove vulnerable individuals from public spaces and reduce their use of crisis services, generating immediate, localized relief⁴.

The Fundamental Solution (Loop B2) requires expanding the overall supply of low-income housing, reforming exclusionary zoning codes, reducing development costs, and strengthening wages and economic protections for low-income households⁶. This pathway requires substantial capital investment and political conflict with property owners, with realization delays spanning years or decades¹⁴.

The Unintended Side Effects and System Erosion (Loop R1) manifest through three primary mechanisms:

System Siltation: Because PSH provides a permanent subsidy with a very low exit rate (12.8% annually), units remain occupied for an average of 7.8 years¹. The PSH stock functions as a permanent capacity sink. Because upstream economic pressures continue to push newly impoverished individuals into homelessness, PSH capacity is quickly exhausted, leaving the system able to serve only 16% of sheltered households¹.
Capital Absorption: Constructing and operating PSH is capital-intensive, with development costs frequently exceeding $500,000 to $800,000 per unit in high-cost urban markets and operating expenses surpassing $20,000 per unit annually¹⁶. These expenses concentrate public funding on a small, high-need cohort, reducing the capital available for broad affordable housing construction that could prevent homelessness upstream⁶.
Political Displacement: By framing PSH as the primary solution to homelessness, political leadership can point to PSH ribbon-cuttings as evidence of progress, diminishing the political will to confront restrictive zoning and real estate market dynamics⁶. As Stroh and Goodman observed, emergency housing investments lower public pressure to resolve systemic housing market failures⁶.

5.2 MOUD and the Pharmacological Containment of Structural Despair

In addiction medicine and public health, the Shifting the Burden archetype appears in the reliance on pharmacological maintenance in place of broader psychosocial recovery infrastructure:

The Problem Symptom consists of fatal opioid overdoses, surging emergency response costs, and rising rates of injection-related infections such as HIV, hepatitis C, and infective endocarditis².

The Symptomatic Solution (Loop B1) centers on the rapid expansion of MOUD (primarily buprenorphine and methadone)²⁵. Pharmacotherapy reduces acute withdrawal, blunts illicit opioid cravings, and decreases all-cause mortality by half and overdose mortality by up to 76% while patients remain in care²⁵.

The Fundamental Solution (Loop B2) involves addressing the root drivers of the opioid crisis: poverty, adverse childhood experiences, untreated psychiatric comorbidities, social isolation, and a lack of recovery capital (such as employment, stable family units, and supportive recovery networks)². Implementing this solution requires integrated community mental health networks, access to psychotherapy, and supportive social environments, all of which require sustained investment and carry long therapeutic delays⁴⁰.

The Unintended Side Effects and System Erosion (Loop R1) emerge across clinical and policy dimensions:

Intervenor Dependence: Prioritizing medication delivery can reduce addiction treatment to biochemical management, sidelining underlying psychiatric conditions and social determinants⁷. The patient becomes biologically reliant on an ongoing medical intervenor⁷.
Attrition Hazards: Because underlying psychological and social drivers are often left unaddressed, patient attrition from MOUD remains high, with median retention times of only 4 to 5 months³. When patients discontinue treatment and lose their pharmacological tolerance, they face a 6-fold surge in all-cause mortality upon re-entering an illicit fentanyl market³⁰.
Atrophy of Comprehensive Treatment Systems: Public resources and regulatory attention are frequently diverted toward low-threshold medication dispensing, leading to underinvestment in intensive residential programs, specialized dual-diagnosis care, cognitive behavioral therapy, and sober living communities¹³. The capacity for fundamental rehabilitation atrophies, leaving the delivery system reliant on indefinite pharmacological maintenance⁷.

5.3 Comparative Systems Archetype Mapping

The structural and systemic parallels between Housing First and MOUD are summarized below:

Structural Archetype Element	Housing First / Permanent Supportive Housing	Medications for Opioid Use Disorder (MOUD)
Problem Symptom	Street encampments, emergency department overcrowding, high municipal crisis and shelter costs	High overdose mortality, surging emergency medical calls, injection-associated viral transmission
Symptomatic Solution (Balancing Loop B1)	Rapid placement into PSH or RRH without preconditions of sobriety or psychiatric compliance	Pharmacological mu-opioid receptor maintenance via methadone or buprenorphine regimens
Symptomatic Mechanism	Spatial/physical buffer: provides physical shelter, mitigates environmental exposure, stabilizes address	Pharmacological buffer: saturates mu-receptors, blocks withdrawal, provides cross-tolerance to overdose
Fundamental Solution (Balancing Loop B2)	Expanding macro-housing supply, zoning reform, increasing affordable housing stocks, income supports	Comprehensive trauma recovery, dual-diagnosis psychiatric care, rebuilding recovery capital and community ties
Fundamental Delays	Multi-year planning, contentious zoning battles, lengthy construction timelines, high capital mobilization	Long-term psychotherapeutic progress, gradual recovery capital accumulation, complex social stabilization
Unintended Side Effect (Reinforcing Loop R1)	PSH siltation (12.8% exit rate, 7.8-year turnover); high capital costs crowd out macro supply; political complacency	Rapid tolerance loss; 6-fold rebound mortality spike post-cessation (RR 6.01); attrition of psychosocial infrastructure
Net Systemic Consequence	Chronic homelessness persists as upstream economic inflows continuously exceed the 16% PSH placement rate	Patients cycle through fragmented care; high discontinuation rates lead to elevated mortality outside treatment

6. Synthesis and Systems-Level Policy Implications

Analyzing Housing First and Medications for Opioid Use Disorder through a system dynamics lens highlights both the essential role and the inherent limitations of harm reduction policies:

First, both interventions operate effectively as operational buffers. Housing First achieves housing retention rates above 84% and drives 24% to 41% reductions in emergency department and hospital utilization⁴. MOUD reduces all-cause mortality by more than half (RR: 0.47) and suppresses overdose mortality (RR: 0.21 to 0.41) while patients are actively maintained on medication³. These interventions provide vital protection against immediate environmental and physiological hazards.

Second, empirical data indicates that neither intervention functions as an internal curative mechanism. Housing First does not produce significant reductions in psychiatric symptom severity or substance use frequency relative to treatment-as-usual comparison groups⁴. MOUD provides no legacy protective effect against fatal overdose once discontinued; rather, treatment cessation triggers an immediate, sharp rebound in mortality risk (RR: 6.01 in the first four weeks), driven by the rapid loss of pharmacological tolerance in an illicit market dominated by synthetic fentanyl³⁰.

Third, both policy areas exhibit clear signs of the Shifting the Burden archetype. Relying on symptomatic buffers without addressing root causes can distort system capacity over time. In the housing sector, low turnover in PSH (12.8% annually) silts up capacity, turning specialized units into static sinks while upstream economic drivers continue to push individuals into homelessness¹. In the addiction sector, prioritizing medication delivery while underfunding broader psychosocial and psychiatric care leaves individuals vulnerable to relapse and fatal overdose whenever treatment is interrupted⁷.

To avoid the pitfalls of the Shifting the Burden dynamic, policy design must avoid treating operational buffers as standalone solutions. Housing First programs must be paired with broader housing market reforms, including zoning liberalization and large-scale affordable housing construction, to maintain throughput across housing stocks⁶. Similarly, MOUD must be integrated into comprehensive, long-term care systems that treat medication as an enabling foundation for trauma recovery, psychiatric treatment, and the development of enduring social capital⁴⁰. Connecting immediate operational buffers to fundamental, long-term interventions is necessary to address the root causes of these crises rather than merely managing their symptoms indoors.

This report was generated by Google Gemini Deep Research using the prompt:

“Research and write a paper on the stocks, flows and system dynamics of

1. Housing first strategies

2. MOUD as a Harm reduction strategy

Your primary goals are

1. To provide national statistics on the stocks and flows and outcomes before and after entering the above stocks (housing first and MOUD) that I could later use to created a “lakes” and “rivers” diagram modeling these

2. to assess whether the impact of each is primary just buffering in terms of successful long-term outcomes:

a. Whether Housing first strategies resulted in significant reductions in substance use severity or mental health symptoms compared to Treatment-as-Usual

b. Whether reduces MOUD reduces all-cause and overdose mortality even after transitioning off the MOUD

3. to assess Whether there are empirical signs that the shifting the burden archetype is active”. It was reviewed by Dr. Andrew Sears.

Works cited

How Much Would It Cost to Provide Housing First to All Households, https://endhomelessness.org/resources/research-and-analysis/how-much-would-it-cost-to-provide-housing-first-to-all-households-staying-in-homeless-shelters/
The cost of addiction: Opioid use disorder in the United States, https://advisory.avalerehealth.com/wp-content/uploads/2025/05/Avalere-Health-White-Paper_The-cost-of-opioid-addiction_OUD-in-the-United-States.pdf
Medication for Opioid Use Disorder After Nonfatal Opioid Overdose, https://www.acpjournals.org/doi/10.7326/M17-3107
Permanent Supportive Housing with Housing First to Reduce … – PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC8513528/
Marc R. Larochelle’s research works | Boston Medical Center, https://www.researchgate.net/scientific-contributions/Marc-R-Larochelle-2064759519
Systems Thinking and Homelessness Policy in Ireland, https://publicpolicy.ie/expert-reflections/systems-thinking-and-homelessness-policy-in-ireland/
Meadows 2008. Thinking in Systems.pdf, https://research.fit.edu/media/site-specific/researchfitedu/coast-climate-adaptation-library/climate-communications/psychology-amp-behavior/Meadows-2008.-Thinking-in-Systems.pdf
Leveraging Grantmaking: Understanding the Dynamics of Complex, https://bridgewaypartners.com/wp-content/uploads/2020/10/leveraging-grantmaking-part-1.pdf
Trainings and Reports — Pinellas Continuum of Care (CoC), https://www.pinellascoc.org/training
The 2023 Annual Homelessness Assessment Report (AHAR to, https://www.huduser.gov/portal/sites/default/files/pdf/2023-AHAR-Part-1.pdf
Inverse systems: national investments in homeless housing, https://www.tandfonline.com/doi/full/10.1080/10530789.2026.2639370
Glossary of terms | Office of Supportive Housing – Santa Clara County, https://osh.santaclaracounty.gov/glossary-terms
Why America’s Homelessness Strategy Failed and How to Fix It, https://ciceroinstitute.org/research/why-americas-homelessness-strategy-failed-and-how-to-fix-it/
SF’s homeless problem: A civic disgrace, https://projects.sfchronicle.com/sf-homeless/civic-disgrace/
At Home/Chez Soi – MAP Centre for Urban Health Solutions, https://maphealth.ca/at-home-chez-soi/
Permanent Supportive Housing as a Solution to Homelessness:, https://ternercenter.berkeley.edu/wp-content/uploads/2026/01/PSH-Paper06.23.pdf
HUD-2026-1453-0001 – Regulations.gov, https://www.regulations.gov/document/HUD-2026-1453-0001
HUD System Performance Measures: Returns to Homelessness, https://mecklenburghousingdata.org/frontpage-article/hud-system-performance-measures-returns-to-homelessness/
Supportive Housing Toolkit – Nevada Housing Coalition, https://nvhousingcoalition.org/resources/supportive-housing-toolkit/
Results From the Canadian At Home/Chez Soi Trial – KMB CAMH, https://kmb.camh.ca/eenet/sites/default/files/2018/Latimer%20et%20al.%2020%20Cost-effectiveness%20of%20HF%20with%20ACT.pdf
At Home/Chez Soi Interim Report, https://www.mentalhealthcommission.ca/wp-content/uploads/drupal/Housing_At_Home_Interim_Report_ENG_0.pdf
Long-term effects of rent supplements and mental health support, https://www.researchgate.net/publication/336342213_Long-term_effects_of_rent_supplements_and_mental_health_support_services_on_housing_and_health_outcomes_of_homeless_adults_with_mental_illness_extension_study_of_the_At_HomeChez_Soi_randomised_control
Medications for Opioid Use Disorder (MOUD) Study – CDC, https://www.cdc.gov/overdose-prevention/data-research/facts-stats/moud-study.html
Opioid Use Disorder Treatment Integrated with Primary Care, https://www.medschool.umaryland.edu/media/som/microsites/DACS/docs/9.27.22-DACS-Training—OUD-in-Primary-Care.pdf
Medication-Assisted Treatment Statistics 2026 – Clear Steps Recovery, https://clearstepsrecovery.com/addiction-recovery-blog/medication-assisted-treatment-statistics
The Effectiveness of Medication-Based Treatment for Opioid Use, https://www.ncbi.nlm.nih.gov/books/NBK541393/
Estimated Reductions in Opioid Overdose Deaths… – Ovid, https://www.ovid.com/02093599-202306000-00001
Methadone and buprenorphine reduce risk of death after opioid, https://www.nih.gov/news-events/news-releases/methadone-buprenorphine-reduce-risk-death-after-opioid-overdose
Treatment with Methadone or Buprenorphine Following Nonfatal, https://www.drugpolicyfacts.org/node/3934
Mortality risk during and after opioid substitution treatment, https://www.uclaisap.org/clinicalproviderquicktips/docs/Arianna-Campbell/Final-Supporting-Materials/Articles/Article-1_0941_Mortality-risk-during-and-after-opioid-substitution-treatment.pdf
Mortality risk during and after opioid substitution treatment, https://findings.org.uk/PHP/dl.php?file=Sordo_L_1_findings.pdf&sf=fpd
Mortality risk during and after opioid substitution treatment – The BMJ, https://www.bmj.com/content/BMJ/357/BMJ.j1550.Full.pdf
Association of Opioid Agonist Treatment With All-Cause Mortality, https://www.researchgate.net/publication/352071293_Association_of_Opioid_Agonist_Treatment_With_All-Cause_Mortality_and_Specific_Causes_of_Death_Among_People_With_Opioid_Dependence_A_Systematic_Review_and_Meta-analysis
Association of Opioid Agonist Treatment With All-Cause Mortality, https://pubmed.ncbi.nlm.nih.gov/34076676/
The At Home/Chez Soi trial protocol: a pragmatic, multi-site … – PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC3221290/
Finding Home – Street Roots, https://www.streetroots.org/finding-home/
Effects of Housing First approaches on health and well-being … – PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC6581117/
Long-term effects of rent supplements and mental health support, https://www.furniturebank.org/?pdfemb-data=eyJ1cmwiOiJodHRwczpcL1wvd3d3LmZ1cm5pdHVyZWJhbmsub3JnXC93cC1jb250ZW50XC91cGxvYWRzXC8yMDE5LUxvbmctdGVybS1lZmZlY3RzLW9mLXJlbnQtc3VwcGxlbWVudHMtYW5kLW1lbnRhbC1oZWFsdGgtc3VwcG9ydC1zZXJ2aWNlcy5wZGYiLCJ0aXRsZSI6IjIwMTktTG9uZy10ZXJtLWVmZmVjdHMtb2YtcmVudC1zdXBwbGVtZW50cy1hbmQtbWVudGFsLWhlYWx0aC1zdXBwb3J0LXNlcnZpY2VzIiwiaW5kZXgiOjEsInBkZklEIjoxMzU3NX0
Effects of comorbid substance use disorders on outcomes in a, https://www.researchgate.net/publication/318100092_Effects_of_comorbid_substance_use_disorders_on_outcomes_in_a_Housing_First_intervention_for_homeless_people_with_mental_illness
findings from the At Home/Chez Soi study, Toronto site – PubMed, https://pubmed.ncbi.nlm.nih.gov/33866385/
Mental and substance use disorders and food insecurity among, https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0232001
The impact of opioid agonist treatment delivered in different settings, https://research-information.bris.ac.uk/en/publications/the-impact-of-opioid-agonist-treatment-delivered-in-different-set/
Unmet need for medication for opioid use disorder among persons, https://stacks.cdc.gov/view/cdc/154709/cdc_154709_DS1.pdf
Neighbors to challenge new supportive housing before supervisors, https://missionlocal.org/2025/07/neighbors-to-challenge-new-permanent-supportive-housing-at-16th-and-mission-before-board-of-supervisors/
(PDF) Homelessness and the American Legal Order: Historical, https://www.researchgate.net/publication/396871666_Homelessness_and_the_American_Legal_Order_Historical_Foundations_Comparative_Perspectives_and_Pathways_to_Reform
How Did We Get Here? Mental Health Accessibility For Adults In, https://digitalcommons.library.uab.edu/cgi/viewcontent.cgi?article=1236&context=etd-2020s
Thinking In Systems | Summary & Notes – Will Patrick, https://www.willpatrick.co.uk/notes/thinking-in-systems-donella-meadows/
Effectiveness of the Housing First model among substance users, https://www.emerald.com/hcs/article/27/2/90/1236347/Effectiveness-of-the-Housing-First-model-among

Metric / Domain	Pre-Entry Baseline or Comparison (TAU)	Post-Entry in Housing First (PSH)	Relative Effect Size / Statistical Significance	Primary Data Source
Homelessness Status	Prolonged street homelessness or continuous emergency shelter reliance	Permanent residential tenure in scattered-site or single-site units	88% reduction vs Treatment First; 89% reduction vs TAU	Peng et al. Systematic Review⁴
Housing Stability (% Days Housed)	30% to 60% of total days stably housed over longitudinal follow-up	80% to 85% of total days stably housed over 2 to 6 years	Adjusted Absolute Gain: +41% to +54%;	At Home/Chez Soi Multi-Site Trial⁴
Emergency Department (ED) Visits	2.5 to 6.0 acute visits per person-year	Statistically significant reduction to 1.8 to 3.5 visits per person-year	24% to 41% relative reduction; rate ratio range: 0.59–0.76	Terner Center; NV Toolkit; Peng et al.⁴
Inpatient Hospitalization Days	Elevated acute psychiatric and medical bed-day utilization	Redirection toward planned outpatient community appointments	29% to 36% decrease in total inpatient days;	Latimer et al.; NV Toolkit; Peng et al.⁴
Incarceration & Police Custody	Frequent misdemeanor arrests, booking, and municipal jail stays	Marked reduction in detentions, loitering, and public nuisance arrests	40% to 50% decrease in acute carceral custody days	Stergiopoulos et al.; MHCC Reports²¹
Direct Public System Cost Offsets	High expenditures across crisis health, shelter, and carceral systems	Net annual savings of $7,910 to $15,000 per housed individual	Direct cost offsets of ~54% on average (up to 80% for high-utilizers)	Latimer et al.; MHCC Economic Analysis²⁰