Working research synopsis — October 2026. This is an analytical research frame, not a completed evaluation. Administrative data, policy interpretation and proposed interventions require validation with the City of Ottawa, service providers and people with lived experience.

Ottawa’s objective of making homelessness brief, sheltered and non-recurring is ultimately a question of system flow. People enter homelessness faster than they leave it; limited exits extend shelter stays; long stays consume capacity needed by new entrants; and congestion pushes the city toward overflow arrangements that cost more while offering less stability. The visible pressure appears at the shelter door, but its causes are distributed across rental affordability, income support, eviction, social-housing allocation, health and support needs, documentation, data systems and the availability of units.

The City’s 2024 Housing Needs Assessment found that shelter inflows consistently exceeded exits to long-term housing. In the 2024 Point-in-Time Count, 49 per cent of respondents were chronically homeless and 30 per cent identified insufficient income for rent and living costs as the principal reason they lost housing. Those figures suggest that Ottawa is not dealing only with emergency accommodation. It is managing the downstream effects of a housing and income system that allows preventable losses and then struggles to restore stability.

The practical research challenge is to distinguish constraints the City cannot quickly change—market rents, provincial benefit levels and the overall supply of deeply affordable housing—from processes it can improve now. Better diversion will not create apartments. Better data integration will not make Ontario Works or ODSP shelter allowances adequate. But earlier intervention, more consistent triage, faster document preparation, more transparent prioritization and fewer hand-offs can reduce avoidable entries and shorten some stays while larger supply and income reforms proceed.

The analytical proposition

The shelter system should be examined as one connected pathway with two principal control points. At the front door, prevention and diversion can stop a temporary financial or family crisis from becoming a shelter entry. At the back door, coordinated access, portable assistance, landlord engagement and housing placement can reduce the time between identifying a viable housing option and moving into it.

The most important measure is therefore not a single occupancy count. It is the movement of people through the system: first-time and repeat entries, the distribution of lengths of stay, the share staying longer than 180 days, time from assessment to match, time from match to move-in, exit destination, and returns to homelessness after six and twelve months. These measures reveal whether a lower shelter count reflects durable housing, temporary displacement or a shift into less visible homelessness.

This approach also changes how success is understood. A diversion is successful only if the person does not return shortly afterward. A housing match is meaningful only if a unit is secured. An exit is durable only if housing is retained. Operational performance should be judged across the full sequence rather than at the point where one agency’s responsibility ends.

Research questions

Primary question

Which changes to diversion, prioritization and housing-placement processes, supported by better data, would most reduce shelter inflow and long shelter stays in Ottawa?

Sub-questions

  1. Front door: What share of people entering shelter could have been diverted through problem-solving conversations, family reunification, landlord mediation or flexible one-time financial assistance? What is the cost per successful diversion, and how many people return within 90 days?
  2. Eviction prevention: Where do evictions associated with shelter entry originate—private rentals, community housing, arrears, N12 landlord-use notices or N13 renovation and demolition notices—and how early can the City intervene through rent-bank applications, landlord engagement or usable Landlord and Tenant Board information?
  3. Long stays: Who remains in shelter for 180 days or more, and which barriers are most consequential: income, unit size, accessibility, support needs, immigration status, credit, references or document readiness?
  4. Prioritization: Does coordinated access and the by-name list match people to suitable housing and supports quickly? Where do delays accumulate between assessment, prioritization, match, offer and move-in?
  5. Systems: How effectively do the shelter information environment, the social-housing registry and provincial systems such as the Social Assistance Management System share data? Where do duplicate entry, inconsistent definitions or missing information slow decisions?
  6. Income adequacy: How do Ontario Works and Ontario Disability Support Program shelter allowances, the Canada–Ontario Housing Benefit and the Canada Disability Benefit interact with Ottawa rents, and what do those relationships mean for exits from shelter?

Policy architecture

The proposed review sits at the intersection of municipal operations and rules made by other orders of government. Ottawa’s 2026–2035 10-Year Housing and Homelessness Plan supplies the central outcomes: prevent housing loss and make homelessness brief, sheltered and non-recurring. The City’s Housing Services Long-Range Financial Plan and Emergency Shelter Standards shape resources and service expectations.

Ontario’s Housing Services Act, 2011 and O. Reg. 367/11 govern important aspects of social-housing access, waiting lists and priority, including the Special Priority Policy for survivors of abuse and trafficking. The Residential Tenancies Act, 2006 structures eviction processes and notice categories, while the Ontario Works Act, 1997 and Ontario Disability Support Program Act, 1997 shape the income available for shelter and basic needs. Ontario’s Homelessness Prevention Program and the Canada–Ontario Housing Benefit provide funding and benefit instruments that can support prevention and exits.

At the federal level, Reaching Home directives require designated communities to use coordinated access, maintain a by-name list and adopt an outcomes-based approach. Those requirements should be treated as operational tools rather than reporting obligations. A real-time by-name list can expose bottlenecks only if definitions are consistent, data are current and agencies can act on what the list shows.

Any integration of shelter, housing and social-assistance information must also comply with the Municipal Freedom of Information and Protection of Privacy Act. The objective should not be indiscriminate data collection. It should be the minimum lawful sharing needed to avoid asking people to repeat the same information, identify delays and coordinate service.

Policy interpretation must be translated into operational decisions. A legislative and directive register should identify each applicable requirement, its effective date, the service decision it governs, the responsible business owner, the affected application or form, and the evidence needed to demonstrate compliance. This is particularly important when provincial benefit rules, tenancy procedures, Reaching Home directives or municipal standards change on different timelines. The test is not whether staff have received a policy circular; it is whether intake scripts, eligibility rules, system fields, notices, escalation paths and quality checks have changed with it.

Research design

The quantitative component would follow shelter entrants from 2022 to 2026, separating first-time from returning users and examining length of stay, chronicity, exits and returns within six and twelve months. Descriptive analysis should establish the long-stay profile; regression can test which observed characteristics are associated with stays exceeding 180 days without treating correlation as an explanation.

Business-process mapping should follow three journeys: diversion at intake, coordinated-access matching and housing placement. Workshops with frontline and system staff can document hand-offs, cycle times and recurrent barriers such as identification, income verification, unit inspection and landlord approval. A systems inventory should identify common fields, data owners, interfaces and duplicate entry across platforms.

Evidence from Calgary, Hamilton, Waterloo Region, Housing First programs and Built for Zero communities can identify plausible practices, but transferability must be tested against Ottawa’s housing market, service network and legal context. Ottawa’s own diversion and housing-based case-management pilots should be examined before importing another jurisdiction’s model.

The research also requires structured participation by people experiencing long shelter stays. Interviews should test whether administrative records capture the barriers people actually face, including discrimination, inaccessible units, family composition, immigration documentation, support intensity and the cumulative effect of repeated rejection.

From policy to operating practice

The work should connect policy analysis to the business applications and procedures through which services are actually delivered. For each priority process, a current-state map should show decision points, hand-offs, wait times, rework and the systems used. A future-state map should then specify which step changes, who holds authority, what information is required and what exception path applies when a routine case becomes complex.

System changes should begin with service requirements rather than software features. A requirements package for the shelter information system, social-housing registry and provincial social-assistance interfaces should define common identifiers, mandatory fields, validation rules, role-based access, audit logs, reporting outputs and retention rules. It should also identify where integration is lawful and useful, where a structured referral is sufficient, and where data should remain separate. Frontline users and people with lived experience should test revised workflows before broad implementation.

Operational guidance should be short enough to use under pressure but precise enough to produce consistent decisions. A standardized diversion protocol, for example, would define screening questions, consent, available flexible funds, supervisor approval thresholds, safety exclusions, documentation and follow-up. It should not impose diversion where emergency shelter is the safe and appropriate response.

Implementation should be managed as an operational change project, not the publication of a report. Each recommendation needs an accountable owner, dependencies, cost, implementation milestone, training requirement, technology change, communications plan and performance measure. A staged rollout can test the approach at selected intake sites, compare results, correct technical and procedural failures, and only then scale. A regular governance table involving municipal program, policy, technology, privacy, legal and service-provider representatives would resolve cross-system issues and control changes after launch.

Evaluation should combine outcome, process and equity measures. A fall in average shelter stay could be positive, or it could conceal exits to unstable arrangements. Faster processing could reduce delay, or simply move the queue to the landlord-approval stage. The evaluation design must therefore pair speed and volume measures with housing retention, repeat entry, client experience and disaggregated outcomes. Findings should feed back into operating procedures on a scheduled cycle rather than wait for the next multi-year plan.

What the research should produce

The final work should deliver a shelter-flow and long-stay profile; current- and future-state process maps with quantified bottlenecks; a legislative-to-operational requirements register; operational guidance for consistent diversion at every intake point; business and data requirements for affected applications; common indicator definitions; and a prioritized implementation plan with costs, responsible institutions and sequencing.

Recommendations should be divided into three categories: changes Ottawa can make administratively; changes requiring new municipal resources or agreements; and reforms dependent on provincial or federal decisions. That separation will keep operational improvements from being used to obscure the central constraint. When rents substantially exceed the income and benefits available to people leaving shelter, process efficiency can improve the system, but it cannot substitute for affordable housing and adequate income.

Indicators for a public dashboard

  • first-time shelter entries and entries following eviction;
  • successful diversions and returns within 90 days;
  • median shelter stay and the number and share of stays exceeding 180 days;
  • assessment-to-match and match-to-move-in times;
  • exits to stable housing and returns within six and twelve months;
  • eviction prevention by intervention type and cost per sustained tenancy; and
  • missing, duplicated or unreconciled records across participating systems.

These indicators should be disaggregated where lawful and useful to identify unequal outcomes, while small-cell suppression and privacy safeguards protect individuals. Publication matters because it turns an aspiration—brief and non-recurring homelessness—into a claim that can be tested over time.

Policy and evidence references for the next phase: City of Ottawa, 2026–2035 10-Year Housing and Homelessness Plan; City of Ottawa, 2024 Housing Needs Assessment and 2024 Point-in-Time Count; Housing Services Act, 2011 and O. Reg. 367/11; Residential Tenancies Act, 2006; Ontario Works Act, 1997; Ontario Disability Support Program Act, 1997; Ontario Homelessness Prevention Program guidelines; Canada–Ontario Housing Benefit; Housing, Infrastructure and Communities Canada, Reaching Home Directives; Municipal Freedom of Information and Protection of Privacy Act; Canadian Observatory on Homelessness, A New Direction: A Framework for Homelessness Prevention.