This article is based on a conversation with Raphaëlle Martínez from the Global Partnership for Education (GPE).
I spoke with Raphaëlle Martínez from GPE, and we found ourselves circling around a deceptively simple question: if everyone agrees education matters, why does it still struggle to win — and keep — serious political and financial commitment?
Raphaëlle comes at this from within a large global education financing partnership, working on education not as a sectoral issue but as a systems challenge embedded in fiscal architecture, development finance, sector economics, and political decision-making, as aid shrinks and domestic budgets tighten. And I come at it from the think tank and evidence-informed policy world — where the problem is rarely the absence of good ideas, and almost always the absence of the conditions that allow good ideas to survive contact with politics, economics, beliefs, etc.
As I reflect on our discussion, it is now clear that we talked about power asymmetries. About who gets to say “yes”, who gets to say “not now”, and who quietly vetoes your beautifully designed reform by deleting a seemingly innocuous budget line – that makes all the difference.
And the uncomfortable conclusion we kept returning to was this: education reform will not scale — and will not sustain — if it remains trapped in the education sector’s own bubble.
The Ministry of Finance is not just another stakeholder: It’s the arena
The trigger of our conversation was an article I wrote about a discussion I hosted for the Land Governance Facility on why and how to work with Ministries of Finance; which was in turn informed by an earlier article addressing how health systems researchers need to re-think their audiences – prompting Raphaëlle to reach out, as many of the dynamics closely mirrored those she was facing in education financing.
She framed the challenge clearly: for years, education has been treated as a sectoral concern — something you reconcile after the “real” economic decisions have been made. Her work involves pushing education upstream: into national development plans, into medium-term expenditure frameworks, and into the financing decisions that determine what is fiscally sustainable before line ministries begin negotiating for scraps.
I agreed, but nudged further. In my experience, ministries of finance aren’t just another stakeholder you need to “engage”. They are often the arena in which sector ambitions are translated into something that can be funded or dismissed.
Over the years, I’ve heard the same story across sectors and countries. A ministry designs a programme; it looks coherent; it is full of technical logic and moral urgency. Then it goes to the Ministry of Finance, and returns thinner. The cuts are rarely ideological in tone. They look “pragmatic”. Communication and branding go first. Monitoring, evaluation and learning next. Public engagement follows. Then any long-term system-strengthening efforts are trimmed down to pilot-sized ambitions.
Not because finance ministries hate education. Often, they don’t. But because they are paid — politically and institutionally — to ask different questions.
“Convincing” finance ministries is not the real problem
This is where Raphaëlle offered an important correction to the usual advocacy narrative. Finance ministries live in a world of trade-offs among competing priorities. and moral appeals about why education is important tend to slide past the decisions they actually have to make.
We shared concerns about elevating sector-level competition into macro-fiscal discussions around revenue and resource consolidation, where it risks distorting public finance and undermining budgetary allocative efficiency.
In many contexts, she argued that ministries of finance do not need to be persuaded that education matters. They already know it. They accept the idea of human capital, productivity, and long-term returns. The barrier is not a lack of belief in education.
The barrier is trust.
What ministries of finance often don’t believe is that the education system will deliver results at the scale implied by the wage bill it consumes. That scepticism reflects real structural constraints rather than misplaced cynicism. Education systems need investment to improve, but they are often asked to prove results before the system capacities required to generate and measure those results are in place.
This resonates with almost every conversation I have had with economic policymakers. They often perceive their ministry, the central bank, and other financial oversight authorities as islands of technical excellence and discipline compared with the sector institutions they are asked to fund.
That changes everything.
Because if the real question is not “why education?” but “will this deliver?”, then the most useful conversation is not a sermon about the returns to schooling; it is a hard, technical, political conversation about delivery, efficiency, impact.
Raphaëlle described the familiar scene: ministers of education talking passionately about curriculum reform or teacher training; ministers of finance replying, essentially: show me outcomes, and show me how you’ll manage the risks. Two languages, spoken fluently, in the same room, but without translation.
And that, in practice, becomes a bottleneck for reform.
The missing layer is not “better translation”, it’s cooperation
Raphaëlle argued that education systems are missing an intermediary layer of people and institutions able to operate between sector realities and fiscal logic.
In the evidence world, we love talking about “translators”, “brokers” and “bridges” — between research and policy, between sectors, between communities. But metaphor matters. These terms imply separation: distinct worlds that never touch unless we build a structure across them, even though, in reality, they are inherently interdependent and cannot function optimally in isolation.
In practice, the “worlds” always overlap. + The problem is that the overlap is thin, informal, and unstructured.
There are people who move between government and think tanks, between policy design and political negotiation. Sometimes they are literally in the same households, the same networks, the same institutions. But often, this movement is institutional and functional rather than cross-sectoral, in a development system that remains highly siloed.
The capacity exists. But cross-sector interactions remain discretionary rather than systematic as incentive structures – budget allocations, accountability frameworks, and career rewards – are sector-specific.
So the challenge is not to invent translation from scratch. It is to identify, strengthen and incentivise the cooperative spaces where that translation can become normal.
Where our respective biases met: Investing in ecosystems, not superheroes
This is where our discussion converged around a shared systems diagnosis and the need to think about broader ecosystems that make better policy stick.
Raphaëlle asked, reasonably, where “think tanks” would fit in this context — and whether they are inevitably ideological, partisan, or Western in form. My short answer was: they vary enormously. And think tanks won’t solve it all, but we agreed they can be a great way into complex policy ecosystems — and one that sector actors often overlook.
A policy space is more likely to produce sensible decisions (or challenge poorly informed ones) when it has density: enough people, institutions, data, publications, debates, and counterarguments circulating that bad decisions get challenged—and good decisions get improved. Density is not just “more evaluation reports”. It’s a thick environment of contestation and learning.
When density is low, and economic policymakers have little or no connections with education policymakers, a ministry of finance can make a decision that quietly undermines education, and nobody notices — because no one sits at that intersection. No one is paid to care. No one has the vocabulary. No one has the incentives.
So rather than betting on a single translator, we should be asking:
- Where are the existing spaces where macroeconomic and sector people already meet (formally or informally)? There are economists working in education who are also members of LACEA and the AERC, where macroeconomists (often past, current and future ministers of finance) are also members.
- Who is missing from those spaces? Are education policy research networks and conferences missing development finance experts, fiscal policymakers, and central bank governors? Do these spaces include experts in how the government works and delivers at different levels?
- What incentives would make education policy and implementation a researchable subject for influential macroeconomists who normally ignore it?
- What incentives would make education researchers take fiscal architecture seriously, not just allocation formulas?
Sometimes the answer is embarrassingly modest: small funds, targeted research commissions, curated working groups, joint convenings with different people in the room. Not glamorous. But sustained and catalytic.
Raphaëlle and I focused our discussion on ministries of finance, but we agreed that other actors like parliaments – where generalists and empiricists rule- equally matter! The civil service and the bureaucratic machinery that is more concerned with how rather than what, and often only concerned with “how things are done here”! Teacher unions that represent the interests of their members first – as they should! Etcetera.
Learn more about evidence use in education
A Peruvian lesson: Power doesn’t always sit where we think it does
To make this concrete, I drew on my Peruvian experience, which I am more familiar with.
Some of the most influential evidence shaping Peru’s education debate and reforms in the first two decades of the 2000s did not all come from education experts. It came from economists with strong connections to the Ministry of Finance, the Central Bank and mainstream economic decision-makers — people whose credibility travelled in those circles. They spoke the language of productivity, labour markets, returns on investments and national competitiveness. That evidence moved.
For instance, research by Pablo Lavado, Joan Martinez and Gustavo Yamada on the quality of higher education and underemployment in Peru helped inform the arguments in favour of a widespread educational reform.
Across other sectors, some of the most powerful drivers for instilling evaluation practice across government were not sectoral reforms at all, but rather ministry of finance mechanisms, such as results-based budgeting requirements that tied programme evaluations to funding. When the Ministry of Finance said “no evaluation, no money”, the entire bureaucracy realised they could collect the data after all.
This doesn’t mean education and other sectoral ministries don’t matter. It means that sector change is often governed by non-sector levers.
Which is precisely why education cannot afford to stay in the education room.
Here is why I think (humbly) On Think Tanks matters. It is one of the only spaces where think tanks working across sectors can come together, develop and strengthen relationships and learn from and to work with each other.
Private capital: Everyone wants education — just not public education
We also spent time on the private finance question.
Raphaëlle noted a blunt reality: impact investors say they want to invest in education, but they usually mean private education — where returns are direct and commercial.
Public education is different. The “returns” are social, political, macroeconomic, and long-term. Investors ask: what’s in it for me?
That question makes many in the education community uncomfortable. It can sound like moral corruption: reducing education to a financial instrument.
But we both think avoidance is a luxury.
If private capital is moving (and it is), then someone needs to shape the policy and regulation frameworks that govern how it moves, where it can enter, and what public safeguards exist. That is not primarily a ministry of education task. It is also often a ministry of finance task: taxes, incentives, regulation, and risk instruments.
And again, this is where think tanks and policy research organisations can matter: not as moral cheerleaders, but as architects of feasible frameworks and as independent evaluators of what those investments actually achieve.
Where shared interests move decisions
At this point, we converged on a principle that sounds obvious and is still routinely violated: decisions move when interests align, not when arguments are simply better framed.
If your counterpart is genuinely interested in “beyond GDP”, then lean into wellbeing, equity, legitimacy, and long-term resilience.
If your counterpart is focused on GDP, productivity, wage bills, and fiscal consolidation, then anchor the discussion using the relevant evidence — without apology.
I think this isn’t selling out. It’s basic political intelligence.
The education sector (as do other sectors) often behaves as if the moral case should be sufficient. It isn’t. Moral cases win applause. Budget cases win appropriations.
The part we avoid: Education reform is a thorny political project
Finally, we landed where many “evidence-informed policy” conversations should land, and too rarely do: on the thorny issues.
I suggested that even if you succeed in aligning education expertise with finance ministry logic, there is still a third layer that can destroy everything: corruption, capture, ideology, vested interests, union politics, illegal economies, and polarisation.
You can design a brilliant programme, fund it properly, and still fail because you pretended these factors were “external risks” rather than core design conditions. Corruption is endemic in Peru (just look at our past presidents languishing in jail or corruption trials). Can anything designed to assume corruption not derail it be expected to work? Can it claim to have been properly designed in the first place?
This is where technocracy breaks.
I offered a harsh Peruvian example: technocratic education reform driven by education sector experts and advocates in the first 15 years of the 2000s made real gains, but it failed to build political constituencies strong enough to defend those gains. As Eduardo Dargent has so clearly demonstrated for Peru, when the technocrats left, the reforms unravelled. Professionalisation reversed. Inclusion agendas were dismantled. The system slid backwards fast.
Policy change does not end. It is contested continuously. If you build a reform as a technical project, it will die as a political project.
Raphaëlle agreed and pointed to a broader pattern in international practice: capacity development support has often prioritised technical and bureaucratic fluency over the political positioning of actors able to engage credibly in contestation, strengthen checks and balances, and broaden the range of voices shaping decisions.
What I took away
I left the conversation thinking less about “how to convince ministries of finance” and more about what it takes to make education reform governable:
- Education must be argued and designed beyond the education bubble.
- Finance ministries respond to delivery credibility, not just moral urgency.
- The missing capacity is built through ecosystems, not heroic individuals.
- Private finance will not wait for the education sector to get comfortable.
- And none of it works unless reforms are designed for politics, not just for implementation.
This discussion ended where many good conversations end: not with a neat solution, but with a clearer map of the problem and a shared suspicion that the next phase of education reform will be won or lost at the intersections.
Not in the sector. In the wider system.