The real toilet question

The simplest way to do this edition of The Big Shift, for instance, would be like this: India has seen a big increase in access to toilets in the last twenty years. The most recent household survey data that we have from the government suggests that the share of Indian households who do not have access to a toilet is now down to 8%.

But that's not the full story.

As my colleague Nileena Suresh writes in this piece for us on the progress of sanitation in India, having access to a toilet is not the same as using it. The indicator that we were looking for was this one: what share of people or households report that they use a toilet, versus the share who do not, meaning that they are still forced to defecate in the open, a practice that has deep repercussions not just for public health, but also for safety and dignity.

We do have that data because the National Family Health Survey's nationally representative household surveys include a set of questions to understand the types of toilet facilities used by household members. It records whether the facilities are shared or exclusively used by a household, and if shared, how many other households share the facility.

This data shows, as Nileena writes in the piece, that "[t]he proportion of people defecating in the open dropped from 70% in 1993 to 19% by 2021, reflecting major progress. However, challenges remain, as about one in four rural households still do not use a toilet facility, highlighting ongoing disparities in access to sanitation."

The trouble is that the most recent NFHS was conducted in 2019-21, and there has definitely been progress in building toilets since then. We know this because administrative data from the government on toilets built under their schemes shows that many more toilets have been built since then. But one of the reasons that we prefer household survey data (also collected and disseminated by the government) is because administrative data does not capture what households are experiencing - whether they actually use those toilets, in this case.

Survey data on access to toilets (from the National Sample Surveys) and on the usage of toilets (from the NFHS) for the same years have in the past shown that the question on access overstates progress on sanitation; in a 2020 National Sample Survey, for example, roughly 15% of households reported not having access to a toilet, while the NFHS survey conducted around the same time found that 20% of households reported that they weren't using toilets.

In the absence of what we really want, which is more recent data on the usage of toilets, we can put together all of the information above to say two things - the share of households without access to toilets is probably now down to under 10%, but the share of households not using a toilet is likely a little higher than that.

A big shift, all the same, but the data, as always, has got to be good, even if it makes things a bit more complicated.