Reconciling Cost-Effectiveness and Cheap Enough

By Tony Senanayake @ 2026-04-17T02:56 (+61)

Two frameworks, one question: whose finances are we talking about, and to what end?

As the development sector gathers at the Skoll World Forum in Oxford this week, against a backdrop of sharp cuts to official development assistance — the US has ended 83% of USAID’s programmes, the UK has cut its aid budget by 40%, and the OECD projects an overall ODA decline of 9–17% in 2025 alone, with projected consequences for millions of lives — two financial frameworks are converging in the sector’s conversations. 

The first is cost-effectiveness: IDinsight has launched a dedicated Cost-Effectiveness Unit and the FCDO’s Best Buys evidence summaries are influencing billions in development spending. The second is the “cheap enough” framework developed by the Mulago Foundation, asking whether solutions are structured to scale through government financing rather than perpetual philanthropy. Both feel urgent right now. And together, they open a broader opportunity: to think more holistically about financial frameworks, not just how to maximise impact per dollar, but what among that which is highly cost-effective is also truly scalable and implementable.

These two frameworks are not in tension. They are sequential, and the sector needs both. But to use them well, it is worth being clear about what each one is actually for.

Tl;dr:

• The most useful framing is to treat cheap enough as a subset of cost-effectiveness: use cost-effectiveness to identify what is worth doing, then use cheap enough to test whether there is a viable path to sustainable financing at scale.
• Cost-effectiveness and cheap enough answer different questions at different points in an intervention’s lifecycle. Conflating them or treating them as a Venn diagram with loose overlap, limits the value of both.
• With ODA shrinking, the opportunity cost of mis-sequencing these frameworks has never been higher.

What this piece proposes:

• Greater investment in willingness-to-pay infrastructure for governments and consumers, drawing on well-developed private sector WTP methodology
• Greater standardisation of cost-effectiveness metrics across studies and sectors, to improve comparability and identify true optimums
• Practitioner skill-building that integrates both frameworks, so organisations can design for impact and financial viability from the outset — because impact comes through implementation, not good intent

Cost-Effectiveness: The Optimisation Tool

Cost-effectiveness is a ratio: total cost divided by total impact. GiveWell’s methodology and the broader Effective Altruism movement have done the most to operationalise this in global development. It enables comparison across unlike interventions on a common scale — typically cost per disability-adjusted life year (DALY) averted in health, or cost per learning-adjusted year of education (LAYE) in education. Constructing the ratio requires explicit choices on both sides.

On the cost side, clarity is needed on whether we are measuring full organisational costs or direct delivery costs only, whether we are using pilot-phase figures or modelled at-scale costs (which can differ substantially), and whose costs are being counted: implementer, government, household, or some combination. On the impact side, the choices are equally consequential: are we drawing on RCT-derived outcomes or real-world delivery at scale; counting only direct first-order outcomes or compounding indirect effects; and which common metric - DALYs in health, LAYEs in education - makes comparison across interventions possible.

Strengths:

• Enables comparison across like interventions on a shared unit of account
• Embeds impact directly into the financial metric
• The strongest tool available for optimising the next marginal dollar of investment and managing opportunity cost
• Applicable not just to philanthropic capital but, in principle, to government and consumer spending too

Limitations:

• Vulnerable to measurement bias: RCT-amenable interventions are systematically favoured over those with diffuse or structural impacts
• Does not tell practitioners what cost to be targeting in the design of interventions
• Says nothing about whether anyone outside the philanthropic world is willing to pay at the required cost level
• A programme can be highly cost-effective and remain permanently donor-dependent

Cheap Enough: The Scaling Reality Check

Where cost-effectiveness is the question a funder asks before writing a cheque, “cheap enough” is the question a practitioner should ask at the drawing board. For some audiences, another way to frame this is simply willingness to pay, and specifically the willingness to pay of the payer at scale. As articulated by the Mulago Foundation’s Kevin Starr, the question is not “is this cost-effective relative to alternatives?” but “is the unit cost within what the system that will ultimately sustain this can realistically absorb?”

The three payers at scale:

• Governments: the dominant payer for most interventions targeting the poorest. Willingness to pay is shaped by fiscal space, competing priorities, and political salience.
• Consumers: the relevant payer in market-based models. Cheap enough is determined by household income and competition for discretionary spend.
• Philanthropy: payer at scale only in narrow cases where the problem is global and the philanthropic market is deep enough to sustain delivery indefinitely.

Strengths:

• Creates a decision-relevant design target that practitioners, implementers, and early-stage funders can align around
• Turns cost from a programme output into a non-negotiable design input
• Forces honesty about the pilot-to-scale transition early, when redesign is still possible
• Anchors analysis in political and fiscal realities, not just technical optimums

Limitations:

• Willingness to pay is genuinely hard to measure, particularly for public goods and social interventions
• The threshold is partly endogenous: a well-resourced organisation with strong advocacy can shift government willingness to pay upward
• Says nothing about impact: cheapness without effectiveness is not a virtue
• Can systematically undervalue high-impact interventions that are structurally expensive

Five Ways to Think About the Relationship

How should the two frameworks relate to each other? There are five ways to think about this, ranging from most to least useful.

1. Cheap Enough ⊂ Cost-Effective — Recommended Approach

Start with the universe of interventions that are sufficiently cost-effective, either at the global optimum or a defensible local optimum. From that universe, ask which are cheap enough for the relevant payer at scale to sustain. Cost-effectiveness sets the evidence threshold; cheap enough tests viability at scale. This sequence seems to reflect how philanthropic capital can work best: determine what is worth doing, then ask whether it can be done sustainably.

2. Cost-Effective ⊂ Cheap Enough

Begin with what the payer at scale can afford, then apply cost-effectiveness analysis within that fiscal envelope. This is a reasonable framing from a government budget perspective, though it risks foreclosing high-impact interventions whose cost structure could be redesigned with the right investment.

3. Partial Overlap — the sector default

Some interventions are cost-effective, some are cheap enough, and some are both. This Venn diagram framing is descriptively accurate, but it is passive — it describes where the sector has arrived rather than how it might navigate more deliberately. The two frameworks end up operating independently rather than in productive sequence.

4. Equivalent — big-aid context only

Cost-effectiveness and cheap enough can converge into the same thing when large-scale philanthropy is the permanent payer at scale. GiveWell-funded interventions like bed net distribution (Against Malaria Foundation) operate close to this model. It is a real and important use case, though not a general framework.

5. Mutually Exclusive — avoid

Treating the two frameworks as operating in entirely separate domains with no relationship to each other seems like the least productive framing of all. Both are fundamentally about the financial conditions required to deliver and sustain impact, and keeping them siloed forecloses the sequencing that makes them useful together.

Implications

The two frameworks are currently converging in practice, driven by two forces: the push to optimise scarce financial resources (the cost-effectiveness moment), and the push to scale through government as payer (the cheap enough imperative). That convergence is healthy, but it also surfaces a real risk. 

An intervention can be implemented at massive scale and still be suboptimal, consuming resources whose opportunity cost, even within the same sector and for the same outcomes, could have driven greater impact elsewhere. Cheap enough without cost-effectiveness risks inefficient scale. The inverse is equally problematic: an intervention can have significant impact potential but no viable path to financing at scale, consigning it to perpetual pilot status. Cost-effectiveness without cheap enough risks perpetual donor dependency.

Used in sequence, the frameworks address this. Cost-effectiveness is the right tool for thinking about opportunity cost and scarce capital, not only from a philanthropic perspective, but increasingly from a government and consumer one too. Once a shortlist of locally optimal interventions has been identified, cheap enough provides the reality check: is there a credible path to sustainable financing at the scale that would actually matter?

The practical implication is that the sector needs investment in both directions simultaneously.

On the cheap enough side:

• Significantly more sophistication is needed in measuring willingness to pay for governments and consumers. The private sector has well-developed WTP methodologies; there are genuine opportunities to build equivalent services for development contexts.
• This means better tools to understand political financing constraints and opportunities: existing budget allocations, fiscal space analysis, procurement processes, and the political economy of government adoption.
• Interventions need to be designed with a specific payer and a specific price point in mind from the outset, not retrofitted to a financing model after the fact.

On the cost-effectiveness side:

• Greater standardisation of the numerator and denominator across studies and sectors would significantly improve comparability and the ability to identify true optimums.
• Practitioner skill-building in cost-effectiveness is as important as funder sophistication. Implementers who understand both the cost-effectiveness landscape and the willingness-to-pay dynamics of their target financing system are better positioned to design interventions that are both impactful and viable.

Ultimately, impact comes through implementation, not good intent. These frameworks are tools in service of that goal, not ends in themselves.