To examine the demographics of the residents who live near the 75 stores, Reuters first identified all U.S. Census block groups within ¼ of a mile of each location in Manhattan and ¾ of a mile in Los Angeles, distances that accounted for the cities' different densities. The demographics of those areas were then calculated using the U.S. Census Bureau's American Community Survey estimates, which cover people polled from 2014 through 2018.

To assess levels of poverty, Reuters categorized each store according to whether the portion of nearby households with income below federal poverty thresholds was higher than 20%, or lower. Ten stores are located in areas where the poverty level was not clearly above or below 20%; six had the technology.

To assess race and ethnic makeup, stores were categorized based on whether nearby residents who identify as white and not Hispanic were the largest group. In one case, it was not possible to determine whether or not the white population was clearly the largest racial group.

The findings in both places, Manhattan and metro Los Angeles, were similar.

Reuters calculated margins of error and checked for statistical significance using standard methods recommended by the Census Bureau.

(Ryan McNeill reported from London; Jeffrey Dastin contributed. Editing by Julie Marquis and Janet Roberts)

By Ryan McNeill