There are multiple definitions/types of racism, but one of the most significant ones to enter discourse and study in the 1960s was institutional racism, which more or less defines policies that are disproportionately being used to affect one race as racist.
My point was that there is a difference between a policy used to negatively impact one race and a policy that happens to negatively impact one race. Statistical anomalies and/or demographic trends are not racist.
That people of color are more likely to be arrested, convicted, or sentenced to prison for drug crimes than white people, despite comparable levels of drug use, is most certainly an example of racism in our society.
"Demographic trends"
There's a rich history of redlining and discriminatory housing practices, which is why communities still tend to be racially segregated. By concentrating police activity on these neighborhoods- perhaps a decision often made today by looking at maps that show rates of arrests- you get a feedback loop or cops arresting people of color in predominately minority neighborhoods, because cops were able to do a bunch of that before. In sum it ends up being institutionally racist, even though there may not have been any intentionally racist decision making throughout the way.
There are of course plenty of solutions, but not arresting lots of people for common harmless activity is a good start. Not focusing all policing around areas with the highest arrest rate would be progress. Figuring out how to desegregate neighborhoods would also be useful. The status quo, however, defaults to racism, which is why institutional racism is so insidious.
https://en.wikipedia.org/wiki/Institutional_racism