health · construction · family: the solution exists but nobody can afford it
the floor asa disease vector
Sealing a dirt floor cuts childhood diarrhea in half — but the only proven fix is cement too many families can't afford
Problem statement
For many households in low-income rural and peri-urban areas, the floor of the home is bare earth — and that dirt floor is a continuous reservoir of fecal pathogens, soil-transmitted helminths (hookworm, roundworm), and protozoa like Giardia. Young children, who crawl, play, and put hands in their mouths at floor level, ingest these pathogens repeatedly, driving diarrhea, anemia, parasite burden, malnutrition, and deficits in physical and cognitive development. The cruel feature of the problem is that the intervention is conceptually simple — seal the floor so it can be cleaned — but the solution with the strongest causal evidence, a cement floor, is not cheap for the households who need it: in Mexico the cement alone was worth about $150 per home, roughly one and a half months' income for the households studied, and the government covered that cost for households whose dirt floor was itself taken as a key indicator of poverty, so the dirt floor persists.
Why this matters
The evidence that flooring is a health intervention, not just a comfort upgrade, is unusually strong. In Mexico's Piso Firme program, which replaced dirt floors with cement at roughly $150 per household, a complete replacement of a home's dirt floors was estimated to produce a ~78% reduction in parasitic infestation, a ~49% drop in childhood diarrhea, an ~81% drop in anemia, and measurable gains in children's cognitive development (the effect of simply being offered the program was smaller: 18.2%, 12.4% and 19.4%) — more cost-effective at improving child cognitive development, the authors find, than Mexico's conditional cash transfer program, and apparently more so than nutritional supplementation. A floor is also durable infrastructure: unlike a course of deworming drugs (which do not act on the protozoa that the Mexican study found still infecting regularly dewormed children), a finished floor is a one-time expenditure that, as the Lancet study notes, needs little sustained behaviour change and is easy to maintain. With most households in the Lancet study's rural Bangladesh and Kenya sites still on unfinished floors (only 9.6% and 5.2% had finished floors at enrolment), closing this gap would prevent an enormous, quiet burden of preventable childhood disease.
What’s been tried and why it hasn’t worked
Cement is the default fix and it works, but it costs money precisely where money is scarcest: in Mexico the government covered the cement (about $150 per home) and households supplied only the labor. Deworming campaigns are limited by frequent reinfection and miss the protozoa that, in the Mexican sample, went with a large incidence of diarrhea and anemia. Pure earthen floors are affordable but, unsealed, are not easy to clean and leave young children ingesting contaminated soil — they don't break the transmission pathway. The real technical constraint is a sealed, cleanable, durable floor at a fraction of cement's cost using locally available materials. Earthen floors sealed with a hardening drying-oil varnish (the approach used by EarthEnable) target exactly this gap, but the sealant chemistry, durability under daily wear and washing, drying time, and quality control of locally trained masons are the binding engineering and supply-chain problems that keep such solutions from scaling.
What would unlock progress
The unlock is a floor system that matches cement's cleanability and durability at a much lower cost and carbon footprint, buildable by locally trained labor from local soil plus a small amount of imported sealant. Advances in bio-based hardening sealants (drying oils, plant resins, bio-polymers), simple field tests for floor durability/cleanability, and installment-financing models that fit the cash flow of poor households would each move the problem. Adjacent precedent exists in low-carbon construction materials and in the way sanitation marketing converted latrines from a subsidized handout into an aspirational purchased product.
Entry points for student teams
A materials team could formulate and bench-test locally sourceable floor sealants (flaxseed/linseed and other drying oils, plant resins, lime stabilization) for hardness, abrasion, water resistance, and cleanability against a cement benchmark. A design/engineering team could develop a cheap field protocol or device to measure whether a finished floor actually resists pathogen-harboring dust and survives repeated wet cleaning. A social-venture team could prototype the installment-financing and mason-training model that makes a sealed floor priced well below cement a viable purchase for a poor family. Relevant disciplines: materials science, civil engineering, public health, social entrepreneurship.
Genes
Structural cousins: stuck the same way, in other fields
Sources
Cattaneo, Galiani, Gertler, Martinez & Titiunik, "Housing, Health, and Happiness," *American Economic Journal: Economic Policy*, 1(1):75–105, 2009 (World Bank Policy Research Working Paper 4214), ; Benjamin-Chung et al., "Household finished flooring and soil-transmitted helminth and Giardia infections among children in rural Bangladesh and Kenya: a prospective cohort study," *The Lancet Global Health*, 2021, )30523-4/fulltext (accessed 2026-06-11) go to source 1 ↗ go to source 2 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
Distinct from `construction-earthen-building-code-exclusion.md`, which addresses structural building codes and seismic engineering for earthen walls; this brief is a health-equity problem about earthen/dirt floors as a disease-transmission vector and the cost barrier to sealed flooring. The Piso Firme AEJ paper (Cattaneo et al. 2009) and the Lancet Global Health flooring cohort study (Benjamin-Chung et al. 2021) are the strongest causal-evidence sources. EarthEnable (Rwanda/Uganda/Kenya) is the live social-venture exemplar; its flax-oil sealant and mason-training model define the practical entry points. Follow-up: durability/lifecycle data on oil-sealed earthen floors vs. cement.
Reconciliation 2026-09-30: Fact repair after reading the whole write-up against both sources it names. The World Bank working paper (Cattaneo, Galiani, Gertler, Martinez and Titiunik, Policy Research Working Paper 4214, April 2007) was fetched in full on 2026-09-30 from https://documents1.worldbank.org/curated/en/857881468288000006/pdf/wps421401update1.pdf (36 pages, read in full). The Lancet Global Health paper (Benjamin-Chung et al. 2021) could not be read at the publisher's page, which returns a robot check; its full text was fetched on 2026-09-30 from Europe PMC (PMC7900607, https://www.ebi.ac.uk/europepmc/webservices/rest/PMC7900607/fullTextXML) and read in full, and its citation record was checked against Crossref (The Lancet Global Health, volume 9, issue 3, pages e301–e308, March 2021). The AEJ citation on the Source line was checked against Crossref (10.1257/pol.1.1.75: volume 1, issue 1, pages 75–105, 2009) and matches. (1) The text said the floors of "hundreds of millions of people" are bare earth. Neither source gives a count of dirt-floor households; the Lancet paper says the infections concerned affect over 1 billion people and that 9.6% of households in its Bangladesh sites and 5.2% in its Kenya sites had finished floors. The sentence now says "many households." (2) The text listed "stunting" among the harms; the Mexican study found no significant difference in height and weight, and the Lancet paper says the infections "can result in anaemia, malnutrition, and deficits in physical and cognitive development." The list now follows the Lancet wording, and "chronic diarrhea" is now "diarrhea." (3) The text called cement "the only rigorously proven solution" that "sits just above the affordability line." Neither source says cement is the only proven floor (the Lancet study grouped wood, tile and concrete together as finished floors) or states an affordability line. The sentence now says cement has the strongest causal evidence and gives what the working paper says: the cement was worth about $150 per home, which the paper equates to about one and one-half months' income, and the program covered that cost for households whose dirt floor was taken as a key indicator of poverty. (4) The 78%, 49% and 81% reductions were presented as what the program "produced." In the paper they are the estimated effect of a complete substitution of dirt floors by cement; the effect of being offered the program was 18.2% fewer parasites, 12.4% less diarrhea and 19.4% less anemia. The sentence now says so. "One of the most cost-effective child-health results ever documented" is not in the paper; it now says what the authors do claim, that the program is more cost-effective at improving child cognitive development than Mexico's conditional cash transfer program and appears to be more cost-effective than nutritional supplementation. (5) The text said "several relevant protozoa resist" deworming drugs and that a sealed floor "keeps working for years." The working paper says the protozoa found in the sample "are not sensitive to albendazole" though the children were regularly dewormed; the Lancet paper says finished floors "require minimal sustained behaviour change and are easy to maintain." Both clauses now say that. (6) "Tens of millions of families still on dirt floors" and the estimate of "70% of households in parts of Uganda" are in neither source and were removed; the sentence now gives the Lancet enrolment figures (9.6% and 5.2% finished floors). (7) Removed because they are in neither source: that concrete is unaffordable to "2+ million families in East Africa," that it is carbon-intensive, cracks without skilled installation, is cold and uncomfortable, and that subsidized floors are left bare or rejected. The sentence now says what the working paper says: Mexico covered the cement (about $150 per home) and households supplied the labor. (8) "Deworming campaigns treat the symptom, not the source" is not in the sources; the sentence now gives the Lancet paper's reason (frequent reinfection) and the working paper's finding of protozoa with a large incidence of diarrhea and anemia. (9) "Dusty" is in neither source and "uncleanable" overstates the working paper's "dirt floors are not easy to clean"; the sentence now says unsealed floors are not easy to clean (working paper) and leave young children ingesting contaminated soil (Lancet paper). (10) "Pioneered by EarthEnable, a 2025 Skoll Award winner": neither source mentions EarthEnable or the award, and no new source was added; the award was removed and "pioneered" is now "used." (11) The entry point's "$60–80 floor" and "a family living on a few dollars a day" are in neither source; the sentence now says a sealed floor priced well below cement, for a poor family. Confirmed in the sources and left as written: roughly $150 per household; 78%, 49% and 81%; gains in cognitive development; hookworm, roundworm (Ascaris) and Giardia as the pathogens; children ingesting soil at floor level; deworming drugs as only partial substitutes; the Source line citations for both papers. Could not be checked, because they are not in either source and no new source is added in this round: EarthEnable's countries, its flax-oil sealant and mason-training model (Source Notes); the drying-oil varnish approach; and the sanitation-marketing analogy for latrines. These are left as written. No title changed. Genome Tags untouched. A second reader went over these repairs on the same day, reading both sources independently (the working paper PDF in full; the Lancet paper from Europe PMC PMC7900607 in full, the publisher page returning a robot check), and reworded items (3), (4) and (9) as given above; the other eight repairs stand as the first reader made them.