Indicators give us a sense, in a single number, how the world is faring on issues too vast to grasp any other way. 

For example, “maternal mortality has fallen by roughly 40% since 2000”; and, “millions more children are vaccinated yearly compared to a generation ago.”

Numbers like these get repeated in speeches, funding appeals, and progress reports, and for good reason. They can rally political attention and money behind a cause.

Global health data reflect the priorities of whoever pays for them. Rarely are those priorities of the people producing or using the data locally.

We tend to take such figures for granted, as though they simply exist, facts fully formed. Yet global health indicators are not a neutral mirror held up to the world. 

Behind every statistic is a chain of choices: what data to collect, using which method, how often, and which model to run those data through before they become the tidy figure in a headline. All of these steps cost money, too, a fact that became all the clearer when the second administration of President Donald Trump shook up the funding landscape. 

Since Trump’s inauguration in January 2025, funding cuts to global health have thrown long-standing measurement systems into crisis and sparked active debate about where to go from here. 

Global health data reflect the priorities of whoever pays for them. Rarely are those priorities of the people producing or using the data locally. It’s time to redirect that balance toward health workers and local leaders. 

The Funders Behind the Numbers

Global health measurement has long depended heavily on a handful of bilateral donors and philanthropic foundations. 

These funders support different types of data sources. For indicators like the number of women treated for postpartum haemorrhage, data would ideally be drawn directly from routine health records that are collected as part of service delivery. But in many low-resource settings, routine health data systems are not yet considered reliable enough. 

Much of what we know about population health in lower- and middle-income countries instead comes from periodic assessments and household surveys, relying on teams of interviewers going door to door, asking about topics like child health, nutrition, and reproductive health. 

Donors have long supported such survey programs. The largest of these is the Demographic and Health Surveys (DHS) Program, running since 1984 with funding from the U.S. Agency for International Development (USAID). Surveys implemented through the DHS represent the single most comprehensive, continuous source of population health data for many countries. 

Philanthropy has played an increasingly large role, too. The Gates Foundation, which was the largest single contributor of development assistance for health worldwide in 2025, strongly advocates for measurable impact. This conviction is reflected in its funding for the Institute for Health Metrics and Evaluation (IHME), a research organization working to generate evidence about the state of health everywhere. 

IHME has quickly become one of the leading sources of global health estimates, a role once held mostly by the World Health Organization and other United Nations  agencies.

Funding in Crisis

When the Trump administration dismantled USAID in early 2025, many of the data systems that had tracked USAID’s programs were also affected. The cuts included funding for the DHS program, causing severe disruption to ongoing and planned surveys across dozens of countries. 

Routine health information systems were hit, too. For example, Kenya’s national health information system went temporarily offline when USAID cuts disrupted payment of a routine service fee.

For decades, global health measurement had mostly changed incrementally, through small tweaks and additions to an already-established system. The events of 2025 presented an abrupt shock. 

Some optimists saw a bright side, hoping the disruption might force change for the better— notably, a shift towards countries taking charge of their own health data systems, instead of depending heavily on external donors. 

So far, that shift did not quite take place. The Gates Foundation stepped in with emergency funding to keep the DHS Program running for at least three years. 

The Foundation’s intervention is a good thing, because the DHS provides an extremely valuable source of population health data. But it also handed more influence to a funder that is already a significant force in shaping what counts as progress in global health. 

Some critical scholars have described the influence of the Gates Foundation as “knowledge philanthropism.” When the Foundation funds data collection while also shaping the concepts and tools used to interpret it, the Foundation deepens its indispensability in knowledge production and its influence over global health governance. 

The Foundation’s expanding role has unfolded alongside an overhaul of United States global health engagement. Since dismantling USAID and withdrawing from the World Health Organization, the U.S. government has rolled out the America First Global Health Strategy. The U.S. is negotiating bilateral agreements that require countries to co-finance their own health programs and meet specific U.S. conditions in exchange for continued support. More than 30 countries had signed on by mid-2026, agreeing to contracts worth over $20 billion in total.

Data systems are explicitly becoming part of funding conditionalities. Such explicit terms make visible something usually hidden: Measurement itself is never neutral, no matter who is paying for it. 

Among the conditions attached to some of these deals is access to countries’ health data systems and pathogen databases. Kenya, Rwanda, and Uganda are among the countries that have agreed to terms along these lines. Pitched as support for these countries’ health sovereignty, the agreements make large-scale funding contingent on giving U.S. agencies’ access to digital health systems and outbreak databases.

Critics note that health ministries negotiating these kinds of deals are structurally outmatched. They run the programs the funding is meant to support but have limited say over how the terms get set. 

In Kenya, the arrangement has already run into resistance. A legal challenge led by the country’s High Court to pause implementation of the agreement is pending judicial review.

Old Dynamics, Repackaged

For African countries like Kenya, the recent agreements reinforce long-standing extractive relations. These power inequalities are an old story in global health. 

Global health charts its origins in early 20th century tropical medicine. It was developed by colonial powers primarily to protect the health of colonizers and to maintain a workforce in colonized territories. 

This pattern of prioritizing colonizers’ interests extended to data, too. Early large-scale health surveys in Africa were designed less to help improve local people’s wellbeing than to track the health of workers useful to colonial economies. 

The same survey-based approach resurfaced decades later, in the 1980s, when structural adjustment programs pushed African governments to shrink public spending. Again, periodic external surveys were favored over sustained investment in routine health data systems. This power dynamic persists today, where people from previously colonizing countries have influence over what gets measured, and why, in previously colonized ones.

Recent events show data-as-politics in its most blatant form. Data systems are explicitly becoming part of funding conditionalities. Such explicit terms make visible something usually hidden: Measurement itself is never neutral, no matter who is paying for it. 

The important question, then, is not whether data are partial, but whose partiality we get. 

Every measurement system reflects certain questions and priorities. This is because, by its very nature, measurement is selective. When we measure, we distill some aspects of reality into numerical figures. Health indicators work something like X-rays: They both clarify and hide things from view, illuminating specific features while casting others into shadow. 

This selective focus is a key power of measurement. It helps make a complex situation legible. But time and resources for measurement are always limited, so choosing to track one thing usually means something else goes unmeasured. The important question, then, is not whether data are partial, but whose partiality we get. 

What’s Lost When Funders Shape Measurement?

When funders shape measurement, one cost relates to relevance. Measurement systems that are primarily set up to meet the information needs of the people funding a program often deprioritize the information needs of the people running that program. 

For instance, a district health manager might need up-to-date information on stockouts or staffing gaps to keep services running day to day. But donor-funded reporting systems are typically built to demonstrate a program’s results to the people paying for it, not to help a manager troubleshoot problems in real time.

Second, funders tend to prefer internationally standardized, comparable reporting in order to show how their investments across different contexts get “bang for the buck.” For example, a facility might be asked to track indicators for a health issue that is not very relevant in the local disease burden, while a condition that is very common goes unreported simply because it is not a priority for funders. That mismatch is demotivating for the health workers and managers who collect the information.

Third, externally-imposed reporting requirements often come with a significant workload. This burden is worsened by fragmentation challenges in global health, where various donor-funded programs are designed and reported on in parallel to one another and to national health information systems. Fragmentation results in the national health system competing for workers’ time and expertise with foreign organizations. 

We might ask whether any funding is better than nothing, or any measurement is better than none at all, especially at the current moment in time.

Walk into low-resource hospital wards across the globe, and you will likely find piles of dog-eared notebooks, registers, and loose sheets of paper sitting alongside newer digital systems in response to demands from various different funders, organizations and levels of government. Put yourself in the shoes of a health worker filling all this in. It is a lot of work, especially considering that most of it might not feel very relevant to you, and you rarely hear back about what the data mean or see them used to change anything locally.

It is therefore worth taking a closer look at the role of funding in shaping who counts, quite literally, when it comes to global health data.

Shifting Purpose and Power

Criticising the very funders keeping global health measurement afloat can feel risky, even self-defeating. We might ask whether any funding is better than nothing, or any measurement is better than none at all, especially at the current moment in time.

There is no clear-cut answer to this challenging question. But even with external funding, time and capacity to collect and use data are limited. Countries can less afford than ever to keep running measurement systems built to satisfy distant funders, especially at the expense of being useful closer to home. Now is the time to take a critical look at whose information needs are being met, and who is accountable to whom. 

Donors have a central role to play in helping countries build cohesive health data ecosystems that prioritize local information needs, over measurement designed mainly for international comparison. What would that require?

First, indicator design should start with what care providers need to know day-to-day, treating their information needs as a primary concern rather than as an afterthought. 

Making that shift requires asking healthcare workers, health facility leaders, and sub-national level managers: What data are relevant for the daily work you do? Indicator design should follow a bottom-up logic, starting from the priorities of those closest to the point of care. 

Second, we should invest in forms of measurement that are locally specific and therefore non-standardizable. Prioritizing locally-specific measurement means resisting the pull toward standardization as the default starting point and committing more time and resources to place-based data that will not travel far. 

Valuing locally-specific measurement does not mean we should stop investing in standardized measurement altogether. Rather, we need to do a better job at distinguishing when standardization is truly required. 

Third, no investment in collecting data should go forward without an equal investment in the local capacity to interpret and use data. Such an approach carries real resource implications, and it challenges the current logic of donor investment which typically directs most funding towards data collection and reporting. More of that funding should instead support local-level data use. 

Country ownership of health data systems is of crucial importance. But ownership does not emerge simply and suddenly by retracting external financial support. 

Data should be used to generate relevant information that fuels learning and  continually improves health systems. Health facilities and data collectors should receive regular updates on data trends, participate in staff meetings and learning exchanges, and share best practices, supported by coaching and mentoring structures that build data-use capacity over time.

The Clinical Information Network in Kenya offers a working model of what using data to enable health system learning can look like in practice. Set up to improve pediatric care, the network strengthened existing medical record systems to inform network priorities. It directly involved health workers in reviewing data to track progress and make improvement plans. This way, the network uses data as a tool for shared learning. Data functions as a means to share experiences and build motivation among care providers. 

Funders should support more initiatives like Kenya’s Clinical Information Network. Money should go toward strengthening national routine health information systems, not building new, parallel ones. Funders should also back the processes behind them: identifying local priorities, building context-specific measurement, and developing the capacity to act on what gets collected.

Rethinking measurement priorities and systems currently in place requires centering the people who actually produce and use health data. These are not philanthropists or donor countries. Rather, health workers, facility teams, and district managers should weigh in on decisions about what gets measured and why. 

Country ownership of health data systems is of crucial importance. But ownership does not emerge simply and suddenly by retracting external financial support. 

Growing local capacity requires sustained investment in the difficult, context-specific work of building systems from the ground up. Funders have long held power to determine priorities in global health. Whether that power will continue to prop up old dependencies in measurement or instead go toward building something more durable is, for now, an open question.