Why don’t households in Bangladesh adopt indoor air purifiers?

A version of this article first appeared in VoxDev.

High levels of ambient air pollution impose severe economic and welfare costs, especially in low-income countries. Over five million people die annually from exposure to ambient air pollution, with South Asia alone accounting for 1.4 million deaths per year. 

Despite persistently high pollution levels, the demand for private preventive technologies like air purifiers remains remarkably low (see Greenstone et al., 2021; Greenstone and Jack, 2015)—even among households that can afford them. For instance, fewer than 1% of middle-class households we contacted in Dhaka, Bangladesh—which is consistently ranked as one of the cities with the worst air pollution in the world—own an air purifier.

Why don’t households that can afford air purifiers, adopt them despite high indoor air pollution?

In our study, we show that households hold incorrect beliefs about the severity of indoor air pollution and the effectiveness of air purifiers—correcting both misconceptions is necessary to increase the adoption and use of air purifiers.

Fact 1: Households have inaccurate beliefs about indoor air pollution.

Using indoor air pollution monitors we installed in households, we show that during the winter months, pollution is staggeringly high, with an hourly average of 133 µg/m³, which is nine times higher than the WHO's 24-hour average guideline of 15 µg/m³. However, while 76% of households believe that the ambient air in their area is severely polluted, they greatly underestimate the severity of indoor pollution, with the same percentage reporting that the air in their homes is only minimally to moderately polluted.

Fact 2: Households have inaccurate beliefs about air purifier effectiveness.

We find that households are uncertain about, and significantly underestimate, the effectiveness of air purifiers in removing indoor pollution. Fewer than half of the households had an opinion on purifier effectiveness, and less than a third believed that purifiers removed more than 25% of air pollution from the rooms where they were used, which contrasts sharply with the actual effectiveness of air purifiers—80% in real-world household environments.

Overall, households have an extremely low willingness to pay (WTP) for air purifiers; the average household WTP was USD 12.2, or 8.4% of the retail cost of air purifiers (USD 150 or BDT 17,000).

Research design to correct misperceptions

To correct these misconceptions, we conducted a multi-phase field experiment providing air monitors and purifiers to households.

Phase 1: Providing indoor air pollution monitors to random households

In November 2023, we recruited 1,008 households from three large housing associations. Eligibility required a functioning WiFi connection and no existing air purifier - criteria met by 99% of interested households. These are middle-income households with an average annual household income above USD 6,000; 34% own an air conditioner, which costs at least twice as much as an air purifier and consumes substantially more electricity.

Immediately following recruitment, we conducted a short Phase 1 survey on perceptions of indoor and outdoor air quality among recruited households.

Then, 512 randomly selected households received an air quality monitor displaying real-time PM2.5 levels along with a chart that categorised these levels from “good” to “hazardous.” These monitors recorded and transmitted minute-by-minute data on indoor PM2.5 levels.

Phase 2: Randomly assigning households indoor air purifiers 

In January 2024, two months later, we conducted the Phase 2 survey, gathering information on household beliefs about outdoor air pollution, indoor air pollution, and air purifiers' benefits. We used a modified Becker-DeGroot-Marschak (BDM) mechanism to elicit households' willingness to pay for air purifiers. Before the elicitation, we informed households that air purifiers remove indoor air pollution.

Subsequently, we randomly provided 345 households with a free air purifier. Each purifier was connected to a WiFi-enabled smart plug, allowing us to collect minute-by-minute usage data. We further randomly assigned these purifier-owning households to receive either no electricity compensation, compensation paid daily, or compensation paid monthly.

Phase 3: Final survey of household perceptions 

In March 2024, we conducted the Phase 3 survey, collecting endline data on perceptions of air purifier benefits, households' willingness to pay for an additional purifier, and their willingness to accept cash to sell back their existing purifiers.

Three key findings:

Result 1: Access to monitors corrects beliefs about indoor air pollution but fails to increase demand for air purifiers.

Providing households with air pollution monitors has a large and statistically significant effect on correcting their perceptions of indoor air pollution and its health risks to adults and children (Figure 1). Households that received monitors were 7.4 percentage points more likely to believe that the air in their homes was severely polluted—a 32% increase compared to households without monitors. However, access to air quality monitors did not increase households' willingness to pay for an air purifier, even though they were informed about its purpose in removing indoor air pollution before elicitation.

Figure 1: Effect of monitors on air pollution beliefs and purifier WTP

Result 2: Experience with purifiers corrects beliefs about their effectiveness but households rarely use them.

Although providing air purifiers corrected misperceptions about their effectiveness (Figure 2), households did not seem to value the devices, as they rarely used them—even when we compensated them for the electricity costs of operating them. Specifically, even with compensation, nearly two out of three households used the air purifier for less than 30 minutes per day (Figure 3).

Figure 2: Effect of purifiers on purifier effectiveness beliefs

Result 3: Access to both monitors and purifiers increases usage and households’ valuation of purifiers.

We show that providing households with both monitors and purifiers increases both air purifier use and the valuation of the purifier relative to households that only receive a purifier (Figure 3). Data from smart plugs indicate that monitors increased purifier use by 270%, equivalent to an additional 74 minutes per day.

Moreover, households that received both monitors and purifiers increased the price they were willing to accept to sell back the air purifiers by BDT 3,800 (USD 32), a 26% increase, compared to households that received only the purifier. 

Lastly, households provided with both a monitor and a purifier increased their WTP between Phase 2 and Phase 3 for a second purifier by 52.7%, although this estimate is underpowered. In contrast, households that received only a purifier saw a much smaller and statistically insignificant 11.7% increase in WTP for a second purifier. Similarly, households that received only a monitor exhibited a small and statistically insignificant 8.1% increase in WTP for the first air purifier.

Figure 3: Effect of monitors and purifiers on purifier usage and valuation 

Why are preventive technologies not used for air pollution?

Our results help reconcile seemingly disparate findings about the under-adoption of preventive health technologies against ambient air pollution. When individuals have accurate beliefs about the effectiveness of a preventive health technology—such as face masks in South Asia—but lack a full understanding of the severity of the underlying health risk, providing information about the risk can lead to increased adoption of the technology (see Baylis et al. 2024, Ahmad et al. 2023). 

Conversely, when people are familiar with the effectiveness of the preventive measure—like air purifiers in China—simply informing them about the severity of the health risk may be sufficient to boost adoption (Ito and Zhang, 2020; Barwick et al., 2024). However, if individuals lack accurate beliefs about both the underlying health risk and the effectiveness of the preventive measure—as is the case with air purifiers in South Asia—correcting beliefs about only one of these factors will not increase adoption.

Therefore, to shift from an equilibrium with incorrect perceptions to one with correct perceptions, a sufficient number of households must change their beliefs about both the health risks and the effectiveness of preventive measures.