Few development interventions seem as straightforward as sanitation. Open defecation contaminates water, spreads pathogens, and kills children. Build latrines, end the contamination, save lives. The logic is clean, intuitive, and backed by decades of epidemiological reasoning.

But when researchers finally subjected this logic to rigorous randomized controlled trials, the results were not what anyone expected. Some of the largest and most carefully designed sanitation studies ever conducted found minimal or zero impact on child health outcomes — the very outcomes the programs were designed to improve.

This doesn't mean sanitation doesn't matter. It means our understanding of how sanitation interventions work — and why they so often don't — requires far more nuance than the simple build-and-benefit model suggests. The evidence tells an important story about the gap between biological plausibility and programmatic reality.

The theoretical case for sanitation is genuinely compelling. Fecal contamination of the environment transmits a wide range of pathogens — bacteria, viruses, protozoa, and helminths — through what epidemiologists call the F-diagram: feces reach new hosts via fluids, fields, flies, fingers, and food. Each pathway represents a route through which a single episode of open defecation can sicken dozens of people.

Diarrheal diseases alone kill roughly 500,000 children under five each year, and inadequate sanitation is implicated in a substantial share of those deaths. Beyond acute illness, repeated enteric infections contribute to environmental enteropathy — chronic gut inflammation that impairs nutrient absorption and is increasingly linked to childhood stunting. The burden extends to soil-transmitted helminths, trachoma, and other neglected tropical diseases.

This disease burden falls disproportionately on the poorest communities. An estimated 673 million people still practice open defecation, concentrated in sub-Saharan Africa and South Asia. The case for intervention seems overwhelming: interrupt fecal-oral transmission at its source, and you should see measurable reductions in child morbidity and mortality.

This reasoning drove massive investment. The Sustainable Development Goals included a specific target to end open defecation by 2030. Governments and donors poured billions into latrine construction programs. But biological plausibility, no matter how strong, is not the same as programmatic effectiveness. The question remained: does building latrines actually produce the health gains the theory predicts?

Takeaway

A strong biological mechanism for how something should work is not evidence that an intervention targeting it will work. Theory tells you where to look; only rigorous evaluation tells you what you've found.

The most definitive evidence comes from a cluster of large-scale randomized controlled trials conducted in the 2010s. The WASH Benefits trials in Bangladesh and Kenya, along with the SHINE trial in Zimbabwe, tested sanitation interventions — often combined with water treatment and hygiene promotion — against control groups receiving no intervention. These were not small pilot studies. They enrolled tens of thousands of households and were designed with enough statistical power to detect meaningful health effects.

The results were striking in their consistency. The WASH Benefits Kenya trial found no significant effect of sanitation on child diarrhea or linear growth. The WASH Benefits Bangladesh trial found modest diarrhea reductions from some combined interventions but no impact from sanitation alone on stunting. The SHINE trial in Zimbabwe — perhaps the most intensive of all, with high-quality latrines and regular follow-up — found no effect on stunting or anemia.

These findings sent shockwaves through the development community. Decades of observational evidence had suggested strong associations between sanitation and child health. But observational studies suffer from confounding — households with better sanitation tend to be wealthier, better educated, and healthier in ways that are difficult to fully control for. When randomization stripped away those confounders, the expected health gains largely disappeared.

It's worth noting what these trials did not show. They did not show that sanitation is biologically irrelevant. They showed that the specific interventions tested, delivered at the coverage levels achieved, did not produce detectable health improvements. That distinction matters enormously for understanding what went wrong and what might work differently.

Takeaway

When rigorous experimental evidence contradicts decades of observational findings, the right response isn't to dismiss either — it's to ask what the experiment reveals about the gap between the intervention as designed and the outcome as hoped.

The most important explanation for these disappointing results is deceptively simple: building a latrine is not the same as eliminating fecal contamination from the environment. Several behavioral and contextual factors explain why hardware alone falls short.

First, adoption and consistent use remain stubbornly difficult. In many trial sites, a significant proportion of household members — particularly men and children — continued defecating in the open even after latrines were installed. Latrines that aren't used by everyone, every time, leave transmission pathways wide open. Second, even in households with full latrine adoption, contamination persists through other channels: children crawling on contaminated soil, animal feces in the domestic environment, and inadequate handwashing after latrine use. The F-diagram has multiple arrows, and blocking just one rarely blocks enough.

Third, and perhaps most critically, sanitation may require community-level saturation to produce individual-level health gains. If your neighbors still practice open defecation, your latrine protects you far less than the theory predicts. The trials typically achieved partial coverage — enough to test the intervention but possibly not enough to cross the epidemiological threshold where transmission is meaningfully interrupted.

This points to a fundamental design challenge. Effective sanitation may require not just infrastructure but sustained behavior change across entire communities, combined with animal waste management, handwashing infrastructure, and water treatment. The intervention that would actually match the biological theory is far more comprehensive — and far more expensive — than what most programs deliver. The question becomes whether development programs can realistically achieve that level of comprehensive change.

Takeaway

When an intervention targets one link in a complex transmission chain, partial adoption doesn't produce partial results — it can produce near-zero results. Thresholds matter, and falling short of them can make the difference between transformation and futility.

The sanitation evidence does not argue against sanitation. It argues against the assumption that building infrastructure automatically translates into health outcomes. The gap between the two is filled with human behavior, community dynamics, and environmental complexity that programs routinely underestimate.

For development practitioners, the lesson extends well beyond latrines. Whenever a program's theory of change relies on a long causal chain — build X, people use X, behavior changes, exposure drops, health improves — every link deserves scrutiny and evidence. Skipping those steps is how good intentions produce disappointing results.

The most productive response to these findings isn't disillusionment. It's better program design: interventions that take behavior as seriously as infrastructure, that target community-level coverage, and that measure what actually changes rather than what gets built.