THE HUMAN RIGHT TO WATER AND SANITATION

A distinguishing feature of the media coverage of the water crises in Flint and Detroit, Michigan has been the linking of those incidents by activists to calls for respect of the human right to water and sanitation (HRWS) (Chapman Reference Chapman2014; Southhall Reference Southall2016). Long fought for by anti-water privatization and other water activists, this right was finally recognized by United Nations General Assembly (UNGA) and the UN Human Rights Council (OHCHR) ^{Footnote 4} in 2010 (Bakker Reference Bakker and Mansfield2008; Barlow Reference Barlow2011; Gleick Reference Gleick1999; MacDonald Reference MacDonald2005). There are three important features of the UN resolutions:

1. 1. the resolutions recognize water and sanitation explicitly as basic human rights, meaning that nation-states and powerful non-state actors have duties and obligations to ensure those rights on the basis of established human rights principles—nondiscrimination being critical in our case;

2. 2. this means that all members of society are entitled to these rights, a change from the development discourse about water and sanitation coverage as a question of proportional targets;

3. 3. the establishment of an OHCHR Special Rapporteur on Water and Sanitation to monitor implementation (see de Albuquerque Reference de Albuquerque2012; Lukacs Reference Lukacs2014; Winkler Reference Winkler2012).

While much of the work of the Special Rapporteur has been in developing countries like Bolivia, Botswana, India, and South Africa, the Special Rapporteur has also reported on wealthy nations, including two fact-finding “missions” to the United States at the invitation of activists and through the intervention of sympathetic U.S. government agents. The first mission (from February 22 - March 4, 2011) resulted in a report that described cases of marginalized communities that lacked access to water and sanitation (de Albuquerque Reference de Albuquerque2011). Based on gathered evidence, the report noted that discrimination on the basis of race associated with lack of access to water and sanitation was a possibility. The second mission, which occurred in July 2014 and was carried out jointly with special rapporteurs on adequate housing and extreme poverty was conducted specifically to investigate massive water shutoffs in Detroit, and resulted in an official statement condemning the water shutoffs (United Nations 2014). Activists, who view the ongoing water crisis in Flint as documentation of potential racial discrimination in water and sanitation services, refer to this report (Lukacs Reference Lukacs2014).

The international acceptance of HRWS as an analytical frame for grievances opens opportunities for consideration of possible discrimination in access to water and sanitation even in places like the United States, where a very small portion of the overall population lack access. However, some scholars, such as Leila Harris et al. (Reference Harris, Rodina and Morinville2015), have argued that the HRWS frame is analytically insufficient and should instead be combined with the conceptual lens of environmental justice to more clearly identify the patterns and mechanisms of discrimination.

RACE AND THE ENVIRONMENT

The scholarship on race and economic inequality has had a long history within sociology. While early theories of race focused on biological differences, by the beginning of the twentieth century, sociologists such as W. E. B. Du Bois (Reference Du Bois1903) and “Chicago School” scholars William Thomas and Florian Znaniecki (Reference Thomas and Zmeniecki1918) started to lay the groundwork for understanding race as a social construction (Winant Reference Winant2000). In its contemporary use, the social construction of race emphasizes systems of domination that produce racialized bodies and racialized politics, resulting in ongoing and everyday reinforcement of systems of domination and the continuation of structures of inequality.

For Joe Feagin and Sean Elias (Reference Feagin and Elias2013), structural racism was foundational to the organization and development of the United States, and remains persistent in its continued maintenance of social inequalities. Structural racism refers to the institutional structures that define the social hierarchies and power of White populations over populations of color. It perpetuates differential access to resources. Whites are explicitly complicit in this social organization, either through direct acts of racism, or through passive and colorblind acts that accept and thereby reinforce their privileged dominance. Scholars such as Douglas Massey (Reference Massey1990), George Lipsitz (Reference Lipsitz2007), and William Rohe and Leslie Stewart (Reference Rohe and Stewart2010) have documented how structural racism around access to mortgage lending, for instance, has led to spatial concentrations of poverty, community vulnerability to economic downturn, neighborhood blight, limitation on intergenerational wealth transfer, and suppressed property values. Structural racism has been shown to impact basic public health and other services to communities of color (Jones Reference Jones2000).

Environmental justice scholars critiqued earlier work on race and the environment for drawing a connection between the two. While many in the environmental movement viewed issues of the environment as not of interest to non-White minorities, environmental justice activists expanded the definition of environment from a space “out there,” far away from people, to “where we live[d], we work[ed], we play[ed]” (Gottlieb Reference Gottlieb2005). As such, early environmental justice scholarship focused on documenting disproportionate exposure to environmental toxins and polluting industries for particular groups (Bullard et al., Reference Bullard, Mohai, Saha and Wright2007). Bullard’s research on environmentally polluting industries in Louisiana and Texas demonstrated that race, rather than socioeconomic status, was the critical issue in determining where these industries and industrial waste plants were located (Bullard Reference Bullard1994).

Bullard’s more recent insight that “the dominant environmental protection paradigm reinforces instead of challenges the stratification of people…places…and types of work” (Bullard Reference Bullard2005, p. 29 ) has sparked new work on the mechanisms through which disproportionate exposure is implemented. Following this lead, new research is focused not only on the relative exposure to environmental risk by minorities and the economically disadvantaged, but also on how White and economically advantaged people systematically avoid exposure to toxics by, for example, moving away from polluted urban centers out to less polluted suburban spaces (Pulido Reference Pulido2000). Laura Pulido’s work further highlights the increasingly recognized spatial inequality in the United States—which, specifically for our purposes can also be described in terms of environmental conditions, protection, and services (e.g., Florida Reference Florida2015; Lobao et al., Reference Lobao, Michael Hooks and Tickamyer2007).

While there is a vast literature on environmental justice that looks at disproportionate exposures to environmental contaminants and the implications for people of color (Brown Reference Brown2007; Pellow Reference Pellow2007; Taylor Reference Taylor2014), there has been less environmental justice work on water and wastewater. What work has been done has tended to treat issues around access to safe drinking water and the avoidance of exposure to the hazards associated with sewerage systems as related, but distinct issues (Wescoat et al., Reference Wescoat, Halvorson, Headington, Replogle, Mutz, Bryner and Kenney2002). Most of the associated research has either focused on community mobilization aimed at redressing unjust siting of “toxic” wastewater treatment plants in already marginalized communities, predominantly women, people of color, and the poor (Di Chiro Reference Di Chiro2004; Litt et al., Reference Litt, Tran, Nga and Burke2002; Pezzoli et al., Reference Pezzoli, Tukey, Sarabia, Zaslavsky, Lynn Miranda, Suk, Lin and Ellisman2007; Ringquist and Clark, Reference Ringquist and Clark1999), or the failure to install or upgrade access to drinking water in a particular political-economic context (Brown and Ingram, Reference Brown and Ingram1987; Wilson et al., Reference Wilson, Heaney, Cooper and Wilson2008).

Some analyses of water and sanitation services access do exist, however. For example, James Wescoat et al. (Reference Wescoat, Headington and Theobald2007) found a connection between water access and poverty in analysis of incomplete plumbing facilities in Colorado. Wescoat et al. (Reference Wescoat, Headington and Theobald2007) not only debunked the myth of universal access to water in the United States, but also demonstrated that low-income water problems occurred beyond “pockets” of impoverishment; instead, scenarios of water and poverty were diverse, and thus deserved diverse solutions. Likewise, Sacoby Wilson et al. (Reference Wilson, Heaney, Cooper and Wilson2008) documented explicit zoning practices in the small town of Mebane, North Carolina that systematically excluded African-American neighborhoods from the development of water and sanitation infrastructure services. Wilson (Reference Wilson2009) then built upon the work of Pulido (Reference Pulido2000) and Juliana Maantay (Reference Maantay2001) in arguing that zoning and structural inequality among places of residence could be correlated not only to exposure to toxins, but also with access to environmental assets, such as access to parks and other green spaces.

Case studies by Jennifer Carrera and Stephen Gasteyer (Reference Carrera and Gasteyer2008) and Carrera (Reference Carrera2014) documented systemic exclusion of Black residents from water and wastewater services in Alabama, Illinois, Mississippi, Michigan, and Ohio. The cases demonstrate processes of structural racism (Bonilla-Silva Reference Bonilla-Silva1997), or what Dennis Rogers and Bruce O’Neil (Reference Rodgers and O’Neill2012) have termed “passive infrastructural violence” (p. 402), the subjugation of non-White populations through the inertia of infrastructure design and planning. Gasteyer and Rahul Vaswani (Reference Gasteyer and Vaswani2004) provided evidence of the extent and variation of lack of access to plumbing through a county-level national analysis. After compiling similar data from the 2000 U.S. Census “long form” (SF3), ^{Footnote 5} they found that racial and ethnic categories, as well as the urban-rural divide, were significant indicators for lack of access. Specifically, rural areas were up to three times as likely to lack complete plumbing facilities, with rural American Indians and Alaskan Native (AIAN) populations representing the greatest percentage of households nationwide that demonstrated a lack of access.

Gregory Pierce and Silvia Jimenez (Reference Pierce and Jimenez2015) used American Housing Survey data to document greater unreliability of water services in mobile homes in the United States. Using county-level data on household expenditures also from the 2000 U.S. Census, Rachel Butts and Gasteyer (Reference Butts and Gasteyer2011) found that the average amount that households paid for water and sanitation in Michigan were higher not only in urban areas, but in areas with greater populations of minority residents. The finding was unexpected, as utility rate theories tended to predict lower water and wastewater service rates in urban areas due to the fixed costs of underlying infrastructure being spread across a larger, and more concentrated population.

Given the unique context of race and racism in the United States (Wacquant Reference Wacquant2008), and using the combined frames of the HRWS, environmental justice, and structural racism, our research therefore asks, do the estimated 540,000 households lacking access to complete plumbing facilities occur in patterns that might indicate discrimination? Based upon this literature, we presume these patterns, should they exist, to be maintained through colorblind laws and legacy conditions that disproportionately increase cost and accessibility of basic infrastructure services for particular places and people.

FINDINGS

Overview

We started our analysis by looking at the population lacking plumbing from a historical perspective. Table 1 adds our ACS 2008–2013 data to a table from Gasteyer and Vaswani (Reference Gasteyer and Vaswani2004) to show the estimated percentage of households lacking access to complete plumbing nationally from 1960 to 2012. From this table, we can see that the United States has improved significantly in terms of aggregate statistics.

Table 1. Percent of Households Lacking Complete Plumbing (1960–2012) ^{Footnote 8}

We found a significant association between lack of access to complete plumbing and region, specifically a positive association with the U.S. Census-defined Pacific, West North Central, and West South Central divisions. Counties which were classified as more rural were also positively correlated with lack of complete plumbing per county. Furthermore, the following social attributes of people were positively correlated with lack of access to complete plumbing by county: percentage of AIAN, percentage of high school graduates, percentage unemployed, and degree of rurality. We found an unexpected weak, but statistically significant, positive relationship with income. Descriptive statistics and both regression results are presented in Table 2 and Table 3. We describe these findings in more depth below.

Table 2. Descriptive Statistics of Socioeconomic Variables

Table 3. Regression of Income, Unemployment, Gender, Race, Age, Education, Urbanicity, and Regional Effects on Lack of Access to Complete Plumbing Facilities in the United States

Regional Effects

The Moran’s I value of 0.016 confirmed that there was indeed clustering within the data, as a significant p-value (p < 0.00) and a positive z-value (z = 15.55) indicated that the observed clustering patterns were not due to random chance. In other words, the raw data demonstrated a tendency for counties with high levels of incomplete plumbing to be clustered near other counties with high levels of incomplete plumbing. Figure 2 shows the percentage of lack of access to complete plumbing facilities by county. Clearly identifiable hot spots include regions in the Southwest, Alaska, the borderlands between the United States and Mexico, and the United States and Canada. This distribution bears a striking similarity to the distribution of hot spots depicted in Gasteyer and Vaswani (Reference Gasteyer and Vaswani2004) (See Figure 1). As mentioned above, when the U.S. Census-defined divisions were included in the model as dummy variables, the Pacific, West North Central, and West South Central divisions produced significant correlations.

Fig. 1. Percentage lacking access to complete plumbing facilities by county (Gasteyer and Vaswani, Reference Gasteyer and Vaswani2004).

Race

Creating a precise update of the 2000 U.S. Census statistics proved significantly challenging, as the U.S. Census Bureau had changed the way it reported race and income statistics in reference to plumbing. It is no longer possible to look at the cross-tabulation of race and access to plumbing directly at the county level. However, Table 4 shows that just by comparing aggregate frequencies by racial group, we can see that while the percentages lacking complete plumbing have fallen across the board, the percentages still vary by racial category: Whites and Asians have the lowest percent lacking complete plumbing, followed by Some Other Race (SOR), Blacks, Hispanic or Latino, and AIAN. What is most striking about this comparison is that the ranking among races has remained remarkably similar.

Table 4. Percent lacking plumbing by race 2000 and 2010 compared

We found a statistically significant relationship between race and plumbing, indicating that counties with a higher percentage of non-White households were more likely to report lack of access to complete plumbing in the 2008–2013 ACS. We performed a second regression to check the results of our first regression; the second regression tested the direct effect of percentage White on the percentage of lack of access to complete plumbing by county. As we anticipated, the second regression showed that the percentage of White per county was significant (p < 0.00, t = -12.38) and negatively correlated with the percentage of households with incomplete plumbing and that for every percentage increase in the White population per county, there was a -0.02% decrease in the likelihood of lack of access to complete plumbing. In other words, counties with larger White populations were more likely to have complete plumbing.

Across all races, we found that AIAN households were by far the population that was most likely to lack complete plumbing by county, followed by Hispanics, Native Hawai’ian Oceanic Pacific Islanders, and Black/African American, although the latter three variables did not produce significant p-values within the regression analyses. Indeed, our regression results show that only AIAN, and SOR populations were significantly associated with lack of complete plumbing per county; however, the direction of their individual effects varied.

The percentage of AIAN households per county produced the strongest effect on percentage of inadequate plumbing; for every percent increase in AIAN households within the county, the percent lacking access to complete plumbing would increase by 10% (p < 0.001, t = 28.54). Conversely, Asian-identified households were negatively correlated with lack of access to complete plumbing facilities: a 1% increase of the Asian population per county generated a 0.04% increase in likelihood of access (p < 0.003, t = -3.01). Therefore, counties with higher Asian populations were more likely to have lower rates of lack of access to complete plumbing facilities. Likewise, as the percent of households that reported being SOR increased in a county, lack of access to complete plumbing facilities per county decreased by 0.04% (p < 0.00, t = -4.30). In other words, the negative correlation with lack of access to complete plumbing also indicated that counties with higher populations of SOR were more likely to enjoy access to plumbing. A county-level spatial analysis likewise revealed that the lack of access to complete plumbing was spatially concentrated in Alaska and the “four corners” (i.e. New Mexico, Arizona, Utah, Nevada) region of the United States—areas with significant AIAN populations, as well as the Texas borderlands region, which are heavily populated by Hispanics and Latinos.