Project Aid in Health: When Earmarked Capital Meets the Realities of Care Delivery

For thirty years, the dominant logic of international health financing has rested on a seemingly straightforward proposition: direct a dedicated pool of capital toward a specific hospital, a particular piece of equipment, or a defined set of clinical services, and health outcomes will follow. This is project aid in its purest form
and it has become the workhorse of global health development. The World Bank, bilateral agencies, and multilateral trusts have channeled billions through project-based mechanisms, each with a project development objective, a results framework, and a closure date.
 

Yet from where I sit—after fourteen years of tracing the arc of medical resource allocation and health insurance economics—the gap between project aid’s theoretical elegance and its operational reality is where the real story unfolds. Project aid offers donors something invaluable: control. Funds are tied to line items, disbursements are linked to milestones, and auditors can trace every dollar from treasury to turbine. But control, in the context of a fragile health system, is a double-edged sword. The very mechanisms that prevent corruption can also prevent adaptation. The infrastructure that gets built may not align with the workforce that is available. The equipment that arrives may require maintenance skills that do not exist locally. And when the project closes, the recurrent costs—salaries, electricity, consumables—often remain, orphaned by a financing model that was never designed for sustainability.
 

This article dissects project aid through a health economist’s lens. I will argue that project aid, when deployed in health sector development, operates as a high-leverage but high-risk instrument. Its effectiveness hinges not on the size of the check but on the alignment between project design and the absorptive capacity of the recipient health system. The core question is not whether project aid works, but under what conditions it works—and for whom.
 

The Control-Sovereignty Tradeoff: Anatomy of a Financing Mechanism

Project aid is defined by its earmarked nature. Unlike budget support, which transfers funds to a recipient government’s general treasury, project aid attaches every dollar to a specific construction target—a road, a water treatment plant, a district hospital. This structure gives the donor country or multilateral agency granular oversight over how funds are spent. The logic is impeccable from a principal-agent perspective: if you want to ensure that a school gets built, you do not hand over a check and hope for the best; you specify the number of classrooms, the type of roofing material, and the contractor selection process, and you inspect each phase before releasing the next tranche.
 

The appeal of this model to health donors is obvious. Health outcomes are difficult to measure in real time; infant mortality declines slowly, and attributing a drop in maternal deaths to a specific aid project is methodologically treacherous. But a completed operating theater is visible. A functioning MRI machine is countable. A trained cohort of community health workers is verifiable. Project aid transforms the amorphous goal of “improving health” into a portfolio of tangible deliverables that can be photographed, audited, and reported to parliament.
 

However, the same mechanism that enables donor control also generates what I call the control-sovereignty tradeoff. When a donor specifies not only what will be built but how it will be built, who will operate it, and what protocols will be followed, the recipient government’s ownership over its own health priorities diminishes. This is not merely a matter of national pride; it has practical consequences for health system functioning. A donor-funded electronic medical records system that does not interface with the national health information platform creates data silos. A vertically funded disease-specific clinic that operates parallel to the primary care network fragments service delivery. A salary top-up for project staff that exceeds government pay scales makes it impossible to retain those workers once the project ends.
 

The literature on aid effectiveness has grappled with this tension for decades. A recent systematic review of development assistance for health found that, despite methodological differences across studies, the weight of evidence indicates a generally positive impact of health aid—particularly in countries with higher governance standards and better economic conditions. This is a crucial qualifier: project aid does not operate in a vacuum. Its success is mediated by the very institutional quality that it often bypasses.
 

The Health Sector Evidence: What the Data Actually Say

If we move beyond theory to empirical evidence, the picture becomes more textured. A study examining project-type interventions found that, while health aid shows a significant and positive effect on health outcomes when interacted with governance variables, project-type interventions themselves showed a significant and relatively strong negative effect on infant mortality in certain specifications. This is a sobering finding. It suggests that the form of aid delivery matters as much as the volume.
 

Why would project aid, which is designed to be targeted and efficient, produce negative effects on something as fundamental as infant survival? The answer lies in what economists call crowding out and opportunity cost. When a project absorbs the best local health administrators into its parallel management structures, it depletes the government’s own capacity. When it introduces new technologies without training sufficient operators, it creates a dependency on external technical assistance that is not sustainable. When it focuses on high-visibility infrastructure in urban centers, it diverts attention—and political capital—away from the less glamorous but more cost-effective primary care interventions that actually drive down child mortality.
 

Conversely, a study on the temporal dynamics of development aid in Africa found that, for the World Bank, positive evidence of aid effectiveness is strongest in the health sector. Health projects, it appears, are more likely to show measurable gains than education or infrastructure projects. This is consistent with the intuition that health interventions—immunization campaigns, malaria control, maternal health services—have shorter causal chains from input to outcome. You distribute bed nets; malaria incidence drops. You train birth attendants; neonatal mortality improves. The signal-to-noise ratio is higher in health than in, say, governance reform or economic policy advice.
 

But even within the health sector, not all project aid is created equal. A quantitative analysis of health aid projects found that 61 percent of projects used a keyword or phrase related to cost-effective interventions, with an average of 11.9 interventions mapped per project. This suggests that donors are increasingly aware of the need to select interventions that offer the greatest health gain per dollar. However, the same analysis found little evidence that this ex ante cost-effectiveness orientation translates into ex post impact measurement. In other words, donors are good at planning for cost-effectiveness but less disciplined about verifying it.
 

The Sustainability Paradox: When Projects End, What Remains?

The most persistent challenge in project aid for health is what I call the sustainability paradox. Projects are designed with a fixed timeline—three years, five years, sometimes seven. They have a beginning, a middle, and an end. But health systems do not operate on project cycles. Diseases do not respect disbursement schedules. Patients do not stop getting sick when the grant expires.
 

Consider the case of the China–World Bank–UK Basic Health Services Project, implemented between 1998 and 2007 to strengthen rural health systems. A retrospective case study found that the integration outcomes of this project varied significantly depending on the level of integration—national versus subnational. Where the project’s interventions were successfully absorbed into domestic health policy, they achieved sustainability. Where they remained as parallel structures, they faded after project closure. The lesson is clear: sustainability is not an automatic byproduct of project aid; it must be deliberately engineered through transition planning, capacity building, and policy alignment from the very first day of project design.
 

The sustainability paradox is exacerbated by the donor incentives that shape project aid. Donors are evaluated by their boards and parliaments on disbursement rates and visible outputs—number of health facilities built, number of health workers trained, number of bed nets distributed. These are metrics that can be reported annually. But sustainability—whether those facilities are still functional five years later, whether those health workers are still employed, whether those bed nets were actually used—is a metric that falls outside the project reporting cycle. The result is a systematic bias toward short-term visibility over long-term viability.
 

From a health economics perspective, this bias is deeply inefficient. The marginal cost of sustaining an existing intervention is almost always lower than the marginal cost of building a new one. Yet project aid, with its emphasis on new construction and new initiatives, consistently underinvests in maintenance, recurrent costs, and institutional strengthening. A hospital that is built but cannot afford electricity is not a health asset; it is a monument to misplaced priorities.
 

Case in Point: Primary Healthcare Infrastructure Under Project Aid

The real-world manifestations of these dynamics are visible across the developing world. In Edo State, Nigeria, a World Bank-supported Immunisation Plus and Malaria Progress by Accelerating Coverage and Transforming Services (IMPACT) Project led to the revitalization and upgrade of 61 primary healthcare centers. This is a classic project aid intervention: targeted capital infusion, defined deliverables, and a clear implementing partner. The immediate results are impressive—61 facilities upgraded, services expanded, coverage increased.
 

But the question that a health economist must ask is not what was built but what will be sustained. Who will pay for the salaries of the additional health workers? Who will cover the cost of vaccines and essential drugs once the project-supplied stock is exhausted? Who will maintain the solar panels and water systems that the project installed? If the answers to these questions are not embedded in the project design—if the recurrent cost burden simply shifts to an already overstretched state government—then the project has not solved the problem; it has merely deferred it.
 

A contrasting example comes from Mali, where a World Bank-supported health center project increased service quality from 37 percent to 81 percent and achieved 87 percent availability of essential drugs. The project constructed 697 new buildings and renovated or renovated 265 others, and patient satisfaction reached 88 percent. These are extraordinary numbers. But they also raise a question: what was the counterfactual? Would the same resources, if channeled through budget support or sector-wide approaches, have produced comparable or better results? We cannot know, because the project was not designed with a rigorous evaluation framework that could isolate its incremental impact.
 

What we do know is that project aid, for all its visibility, is not always the most efficient mechanism for strengthening health systems. Core funding to multilateral agencies—unearmarked, flexible, and aligned with recipient country priorities—appears to be more effective than earmarked funding with regard to outcome effectiveness at the project level. This is a finding that should give pause to any donor who insists on tying every dollar to a specific line item.
 

Rethinking the Project Aid Model: Toward Adaptive, System-Aligned Financing

If project aid is here to stay—and it is, given the political and bureaucratic realities of donor financing—then the task is not to abandon it but to reform it. Based on my analysis of medical resource allocation and health insurance economics, I propose three core principles for more effective project aid in the health sector.
 

First, projects must be designed with exit strategies that are credible and funded. Sustainability cannot be an afterthought. Every project should have a transition plan that specifies, from day one, how recurrent costs will be covered after project closure, how trained staff will be absorbed into the government payroll, and how equipment will be maintained. This plan should be costed and, where possible, pre-funded through an escrow account or a government budget line.
 

Second, project aid must be aligned with national health strategies and systems. The proliferation of donor-funded vertical programs—HIV, malaria, tuberculosis, maternal health—has created fragmentation that undermines the very health systems that these programs claim to strengthen. Project aid should be designed to fill gaps in the national health plan, not to create parallel structures that operate outside it. This requires donors to cede some control—to accept that their funds will be spent according to recipient country priorities, not donor checklists.
 

Third, project aid must be evaluated with the same rigor that donors demand of the projects they fund. The absence of credible impact evaluations in project aid is a scandal. If we do not know whether a project actually improved health outcomes—and by how much, and at what cost—then we cannot say whether it was a good investment. Donors should allocate a minimum of 5 percent of project budgets to independent evaluation, and they should publish the results, regardless of whether they are flattering.
 

Conclusion: The Instrument Is Not the Outcome

Project aid is a tool, not a strategy. It is a mechanism for transferring resources from donors to recipients, but it is not, by itself, a theory of change for health system strengthening. The evidence suggests that project aid can work—particularly in the health sector, where interventions have short causal chains and measurable outcomes. But it works best when it is aligned with recipient country systems, when it is designed with sustainability in mind, and when it is subjected to rigorous evaluation.
 

The control that project aid offers donors is both its greatest strength and its greatest weakness. It prevents leakage, but it also prevents adaptation. It ensures accountability, but it can also undermine ownership. The challenge for health economists and policymakers is to strike the right balance—to design projects that are focused enough to be effective but flexible enough to be responsive, and to fund them in ways that build systems rather than bypass them.
 

After fourteen years in this field, I have learned that the most expensive mistake in health financing is not spending too much; it is spending poorly. Project aid, at its best, can be a powerful instrument for expanding access to care, building infrastructure, and saving lives. At its worst, it can be a costly distraction that leaves health systems weaker than they were before. The difference between these two outcomes is not determined by the size of the budget but by the quality of the design, the depth of the partnership, and the courage to ask the hard questions—not just about what we are building, but about what we are leaving behind.
 

Reference Block
 

Source Reference Link: https://wiki.mbalib.com/wiki/%E9%A1%B9%E7%9B%AE%E6%8F%B4%E5%8A%A9
 

Link Brief: This MBA智库百科 entry defines project aid as earmarked assistance for specific construction projects such as roads, water facilities, and schools. It outlines the mechanism’s core characteristics—donor control, potential for paternalism, and efficiency advantages—which serve as the conceptual foundation for this article’s health-sector analysis.
 

Content Disclaimer
 

This article is for general reference only and does not constitute professional R&D guidance, production process advice or quality certification. All material performance data has specific test premises; readers should verify parameters against actual equipment and working conditions.