logoalt Hacker News

arjietoday at 6:57 PM0 repliesview on HN

To put actual numbers to "way more likely". Using https://www.bayareaequityatlas.org/indicators/experiencing-h... and the census.

The rate is 42/10k for non-Hispanic White and 56/10k for all others - a common definition for people of color. So about a third more likely.

To provide some intuition if the numbers aren't clear that increased risk is about the same as:

* driving while having slept 6-7 hours rather than 7+ hours[0] 1.3x

* being a teen driver with a teen friend in the car rather than alone[1] 1.49x

* the increased cardiovascular risk from being highly sedentary vs. being active[2] 1.3x

As a whole, though, 'person of color' seems to be a bad grouping unit here because Asian people are 'people of color' and have a 7/10k rate, while Black people are also 'people of color' and have a 214/10k rate. Therefore, if I were trying to understand if there are subpopulation differences this is not the aggregation unit I would use.

0: https://www.cdc.gov/niosh/bulletin/2017/national-police-week...

1: https://pubmed.ncbi.nlm.nih.gov/10735394/

2: https://pubmed.ncbi.nlm.nih.gov/38081421/