Or statistical noise. This is exactly what you would expect to see if the data were sliced up and compared using a loose significance threshold like 'p<0.05' in

Or statistical noise. This is exactly what you would expect to see if the data were sliced up and compared using a loose significance threshold like 'p<0.05' instead of a tight one like 'p<(0.05/numberOfZipcodes)'. They don't seem to specify this? See also "Justify your alpha" https://www.nature.com/articles/s41562-018-0311-x https://www.nature.com/articles/s41562-018-0311-x