Contradictory siting studies differed mainly over geographic unit and comparison scope

Evidence

Researchers at the Social and Demographic Research Institute (SADRI) of the University of Massachusetts challenged the findings of the 1987 CRJ study. The SADRI study found that there was not a statistically significant pattern of racial or ethnic disparity in the distribution of commercial hazardous waste sites. Though the SADRI study analyzed data similar to those used in the recent CRJ study, its methodology was significantly different. For example, the SADRI study used census tracts from 1980 and 1990, instead of zip codes, as the geographic unit of analysis. The SADRI researchers also used data from only metropolitan or rural counties, not from the entire United States, as their comparison group (nonhost tracts), possibly understating the relationship among race, ethnicity, and siting choices. Moreover, the study looked only at African American and Latino populations and did not measure the proximity of other racial groups, such as Asians and Native Americans, to existing toxic waste sites. Hence, the researchers’ conclusions leave out a not-insignificant percentage of the people of color population in the United States and possibly understate racial and ethnic disparities. — Luke W. Cole and Sheila R. Foster, From the Ground Up, 300

Categorization

• SEED: Angles toward an essay on how geographic ‘units of analysis’ choices can manufacture or erase evidence of structural racism.
• XREF: Pairs with the earlier claims in ch1 about single-bad-actor frames and in ch3 about documentation, plus general epistemology threads. — connects to: Disparate-racial-impact showings can be defeated by comparing alternative routes’ impacts — Luke W. Cole and Sheila R. Foster, From the Ground Up: Environmental…, 1127; Changing the geographic unit of analysis can make racial siting disparities statistically vanish; Inconsistent research findings often signal underlying measurement error rather than real variation — Bob Altemeyer, The Authoritarian Specter, 2057

Broader Connections

• Inconsistent research findings often signal underlying measurement error rather than real variation #qualifies (0.80) — Bob Altemeyer, The Authoritarian Specter, 2057