I recommend reading the WaPo article that goes along with it: https://www.washingtonpost.com/business/2023/09/28/minoritie...

Bloomberg's choosing to misrepresent the data here - this is not about jobs added, it's about changes in the employment composition.

Simple example: Company X has 950 white and 50 POC employes. 10% leave over the year (95 white, 5 POC). They hire 200 more at an even split (50% white, 50% POC). They now have 1100 people, 955 white, 145 POC. So they've gained net 100 folks - and the net change is +5 white, +95 POC. Voila, 95% people of color hired.

It's still a pretty stunning change with a large ramp up in hiring of POC, but it's much less an indicator of preferential hiring than the Bloomberg framing makes it sound.