That air pollution impacts health is hardly news. Previous studies attest to the fact that air quality, along with other health indicators, further feed into – and even explain – socioeconomic inequalities that already exist. Studies in United States have shown that minority-poor neighbors experience higher pollution levels than White-nonpoor ones, which reflects in their health outcomes. Not only that, Sharkey et al. (2014) show how most of these ‘environmentally disadvantaged’ areas tend to be clustered and ghettoised together, alienating them even further from areas with a better air quality or simply more economic opportunities.
But, invariably, most studies mapping the effects of air pollution on an individual’s health in terms of the neighborhood s/he lives in, and the adjoining areas. Therein lies a catch: individuals living in an environmentally disadvantaged area that has high air pollution levels do not spend all their time there. Recent studies show that they may well spend time beyond their residential areas and the adjacent ones, often traveling to distant areas for work or leisure. If the area they frequent for work or leisure also has high air pollution levels, it reduces the impact living in an environmentally disadvantaged neighborhood. On the other hand, if that area has high air pollution levels, it only furthers it further.
Moreover, one also needs to realise that these movements, although individual, are not necessarily individual driven. In other words, these networks and flows are dependent not only on individual choices but by institutional and social connections. Therefore, these flows are well document at the level of a community than just at the level of an individual.
Noli Brazil, a human ecologist from the University of California, Davis, sought to address this question by tapping into anonymised cellphone data on the movement of a person along with air quality indices. The cellphone data here was provided by SafeGraph, an organization that monitors and maintains a compendium of geospatial datasets for more than forty million American smartphones. This is used as a proxy for urban mobility patterns at an individual level for 88 most populated American cities. This was supplemented by air quality data provided by Environment Protection Agency (EPA) on particulate matter concentration (PM 2.5). The parameter measures particles smaller than 2.5 µm in diameter that can be inhaled in terms of µg/m³. Finally, the resultant datapoints were examined from the standpoint of income levels (poor v nonpoor) and race (White, Black, Hispanic, Asian etc.).
This exercise was conducted at three spatial levels: the residential neighborhood, the neighborhoods adjacent to the residence and the neighborhoods travelled-to for work/leisure/social commitments. These are labeled as the ‘residential,’ ‘adjacent,’ and the ‘network’ levels.
Brazil found that, ‘on average the neighborhoods that residents from non-White communities travel to have higher PM2.5 levels than the neighborhoods connected to White communities. The PM2.5 levels in Hispanic, Black, and Asian networks are 12.4%, 11.5%, and 11.5% higher than the levels in White networks (7.81), respectively.’ Similar results were found in terms of income levels, in that the areas where people from poorer neighbors commute to have 6.8% higher PM2.5 levels than areas visited by people from nonpoor neighbors.
Corroborating prior epidemiological studies, results not only indicate that neighborhoods where Hispanic, Black and Asian groups live are consistently marked by higher PM 2.5 levels compared to white neighborhoods. While PM2.5 levels in white neighbors were 7.81, those in Hispanic, Black and Asian neighbors were 8.85, 8.72 and 8.74 consistently. Furthermore, and predictably so, these neighbors are surrounded by those with similar PM2.5 levels. Bringing to fore disparities across race/ethnicity and income-level groups, Brazil finds that average PM2.5 levels ‘5.9, 5.9, and 5.2% lower in White nonpoor residential, adjacent, and network neighborhoods, respectively, than in poor neighborhoods.’ The advantage of nonpoor neighbors over poor ones, in terms of air pollution risk, was far less pronounced for other ethnic groups like Hispanics, Blacks and Asians.
Out of the three spatial/ecological levels examined here, the third, ‘network’ level, becomes significant not only because it is seldom considered in most studies on urban mobility, but also because – as this study finds – it turns out that people from Black, Hispanic and Asian neighbors travel to distant areas just as much as their counterparts from White neighbors. In fact, Black people travel even further distances than White people do, while commuting distances between Asians, Hispanics and White people are somewhat similar. Examining the ‘network’ level reveals more nuanced patterns as well. For instance, while Hispanic neighbors – poor or nonpoor – carry the most exposure risk in terms of air pollution, but their residents travel to areas that have lower PM2.5 levels. This is not true for Black or Asian groups, which commute to areas that have an exposure risk similar-to or higher-than their residential neighborhoods; thereby making their disparities with White poor/nonpoor groups more pronounced.
The key takeaway that this exercise gives us is that disparities in access to clean air, in terms of both ethnic background as well as the level, extend well beyond the places where people live and that their ‘networks’ play a major role. If anything, its influence had been ‘underestimated’ in previous studies. Brazil hopes that ‘Adopting a network perspective can also increase efficiency in resource allocation by focusing interventions in the most polluted and visited neighborhoods within a mobility network.’