Canada’s Public Health Agency tracked 1,700 communities supported through its Innovation Strategy program. 82% of projects continued after funding ended, and 90% reported sustained impact on public health policies or practices. The results suggest that multi-year financing was part of a multifactorial model; they do not demonstrate that contract duration is the determining variable.

The Essentials

  • Long-term financing, which also covers learning and adjustments, produces programs that persist after subsidy ends.
  • 82% of Canada’s Innovation Strategy projects continued after funding ended; 90% reported sustained impact on public health policies or practices (Public Health Agency of Canada, program active in over 1,700 communities).
  • The central mechanism: enabling project leaders to correct course, build partnerships, and document what works before scaling up.
  • The real tension opposes two financing logics: paying for service delivery or paying for an organization’s capacity to learn.
  • The challenge for 2026-2040 is to anchor this logic without transforming it into permanent funding for organizations that have lost momentum.

Duration as a Design Variable

When a public funder finances a social program, it typically purchases a service: workshops, consultations, support, a number of beneficiaries reached. The horizon is one to three years. After that, the report is submitted, funding stops, and the program often disappears with it. This model rests on a rarely articulated assumption: that the project leader knows exactly what they are doing from day one, that the initial design is correct, and that adjustment is a form of failure rather than a form of work.

This assumption is rarely true. Social innovations operate in fragmented contexts, with populations that aggregate data describe poorly. An intervention designed for families facing food insecurity in a suburban Montreal neighborhood does not translate directly to a similar neighborhood in Winnipeg. Local actors, networks of trust, and relationships with institutions vary. A program that lacks the time to observe these gaps, understand them, and adapt to them delivers a standardized service to situations that are not standardized.

Canada’s Innovation Strategy combined long-term phased financing, adaptation, evaluation, selection, and partnerships to promote quality interventions and their scaling up. Figures published by the Public Health Agency of Canada report a significant portion of projects continuing after subsidy ended and declaring influence on practices or policies.

The Limitations of Short Cycles

The United States offers a useful point of comparison, not because its model would be faulty, but because its institutional constraints are documented and reveal what duration makes possible or impossible.

U.S. federal funding for community health programs often follows annual or biennial cycles, subject to the vagaries of parliamentary appropriations. This structure is not chosen by ideology; it results from the mechanics of the federal budget and the multiplicity of agencies involved. Its effects are known to practitioners: organizations spend a significant portion of their time writing renewal applications, adjusting their mission to changing funder priorities, and maintaining staff whose positions are not guaranteed beyond the current cycle. In this environment, investing in documenting practices or building local partnerships becomes rational only in the very short term. Organizational learning, however, takes time.

Work from the field of social policy evaluation, particularly Paul Osterman’s research on job quality in nonprofit organizations, shows that financing instability translates directly into team instability and discontinuity in relationships with beneficiaries. Yet continuity of relationships is precisely what allows a program to reach populations most distant from institutional systems.

Organizational Learning as Infrastructure

Partnerships were identified as a sustainability factor by the evaluation, but their precise causal effect was not quantified. A program that lasts long enough can also document what works and adjust its methods before scaling up.

It can produce data on its results, which facilitates access to other funding sources. It can train people who, once the program ends, become relays in other organizations. And most importantly, it can transform its practices into local norms, or even institutional policies.

The 90% figure concerns influence or change in public health policies and practices; it does not precisely measure diffusion. A program modifies the environment in which it operates: a practice integrated into a hospital protocol, an approach adopted by a municipality, a tool taken up by another actor. These effects remain invisible in a short cycle, where the program ends before it can create these connections.

Canada’s Mental Health Promotion Innovation Fund, another Public Health Agency program, rests on a similar logic: financing not just activities, but organizations’ capacity to document what they learn and share it. This orientation toward producing usable knowledge is what distinguishes financing for learning from a simple service contract.

It should be noted here that these results are self-reported. Self-evaluation has known limitations: organizations have an interest in reporting successes to their funders. Independent impact studies with control groups would allow verification of the solidity of these figures. Canada is not exempt from bias in its own evaluation.

Project Leaders’ Practices Evolve Over Time

A funder that structures long-term financing changes the incentives of the organizations it supports. A project leader who knows they will have four to five years can afford to recruit and train qualified personnel, knowing that this investment will be amortized over the duration. They can dedicate time to relationships with other local actors without this time being perceived as waste. They can fail, correct, fail again differently, and finally find what works at the precise moment.

This freedom to iterate is what distinguishes social innovation from social service delivery. A service is executed according to a defined protocol. An innovation seeks what works better. Paying organizations to innovate with cycles adapted to service delivery amounts to asking them to improvise within the timeframe of a standard contract.

Mariana Mazzucato’s work on the entrepreneurial state documents comparable logic in research and development: innovations that scale up are often those that benefited from patient financing, capable of absorbing iterations and pivots. Community health is not basic research, but the principle applies: uncertain and long-term results cannot be financed with instruments designed for certain and immediate results. This question of patient financing also connects to the broader debate about what budgetary constraints do to the state’s capacity for action.

Financing Learning Without Creating Permanent Bureaucracies

Herein lies the real tension of this model, and it deserves to be posed directly.

Long-term financing can produce two radically different trajectories. In the Canadian program, continuity was notably associated with other funding sources or integration into existing systems; the precise mechanisms varied by project. This is the scenario that Canada’s 82% continuity appears to illustrate. In the second, the organization settles into the long term, adjusts its practices to satisfy funder requirements rather than field needs, and ends up structurally dependent on public financing without having produced lasting transformation. Long-term financing can create dependence as well as autonomy.

Financing design can be a factor among others; the sources do not allow making it the determinative explanation. A program that finances learning must define from the outset what it means by that: which organizational competencies are targeted, what types of partnerships are expected, what milestones verify that capacity is being built rather than stagnating. Without this architecture, duration alone produces nothing; it simply prolongs what already existed.

The question posed at the 2026-2040 horizon is precisely that one. As governments compare results from community health programs financed differently, pressure to adopt longer cycles will likely grow. The temptation will be to lengthen contracts without changing the piloting logic: paying for activities over a longer period rather than financing capacity to learn. This degraded version of long-term financing might not produce the results observed in Canada.

A signal to watch: the way funders structure their indicators. A program piloted by activity indicators, even over five years, remains a service delivery program. A program piloted by indicators of organizational capacity, partnerships built, practices diffused in the local environment, begins to resemble what the Canadian model measures. To reproduce the model, the PHAC identifies several joint components, including financing duration and flexibility, evaluation, partnerships, systemic change, and knowledge sharing.

The question of evaluation is central in this regard. Programs like Canada’s Innovation Strategy produce self-evaluation data. For a long-term financing policy to be politically defensible, faced with elected officials who prefer short-term results, it must rely on independent evaluations with rigorous protocols. Without this robust evidence base, patient financing remains vulnerable to the political cycles it seeks precisely to circumvent. The legitimacy of a social program depends as much on its capacity to demonstrate its results as on the results themselves, a finding that also applies to digital innovations in public services, as experiences with AI in administrations have shown.

The Structural Decision of Funders

Public funders looking at Canadian results have a concrete decision to make. Paying for service delivery amounts to purchasing a product whose characteristics one controls. Paying for an organization’s capacity to learn amounts to investing in an asset whose value is realized outside the direct perimeter of the contract. This second logic is more demanding to pilot, less transparent for financial departments, and politically more difficult to defend during periods of budgetary constraint.

In this program, a multifactorial financing model coexisted with 82% of projects still active after phase 3; this observation does not measure a causal association specific to financing.

Short-term financing can be very generous and leave nothing behind. The Innovation Strategy had approximately 6.2 million Canadian dollars per year on average and reported influences on policies and practices; the extent of a sustainable sectoral transformation is not established. Financing design is one factor among others in organizations’ capacity to act and adapt. Financing design choices can influence architecture and indicators, but persistence after subsidy depends on several factors that are not isolated causally.

Several governments today seek to produce more results with constrained budgets, which makes this design choice directly operational.

Sources

  1. Public Health Agency of Canada, Innovation Strategy
  2. Public Health Agency of Canada, Mental Health Promotion Innovation Fund (Public Health Agency of Canada, program page)
  3. Paul Osterman, work on job quality in nonprofit organizations and local labor markets
  4. Mariana Mazzucato, The Entrepreneurial State (Anthem Press, 2013; Penguin edition, 2018)