To clarify, it's not so much a Harvard study as it is a working paper by a Harvard professor for the National Bureau of Economic Research. Its conclusions seem less definitive and more directional, self-admitting that its data base might be flawed but that using best practices, this is what it found with what it had.
So what did it have and what did it find?
With regard to shootings, it looked at about 4% of the population in areas serving more black people than average. It looked at only information made available by those police departments (and not by, say, a police department which might have had stats indicating otherwise) and then it looked only at police reports of incidents and not at what witnesses saw or what citizens involved saw or felt for that matter. Also, with regard to lethal incidents, it looked only at death by gunshot and not other deaths occurring while within police custody.
The paper does not deny that a plurality of black men die by the hand of police (whether or not those deaths might be justified). Nor does it deny--but rather reaffirms--that in non-lethal incidents, more force is experienced by blacks and Hispanics than by whites at the hands of police. What is the surprising find of the study which ought not be removed from the light of these considerations is not that the study debunks any shooting myth as the title of this thread presupposes, but rather an oddity that while use of force is found to be racially determined in nonlethal actions, by the methods of study and the questionable data, there was determined to be no racial motivator in applying lethal action with regard to bullets.
So, again, it doesn't say blacks won't be pulled over more in their cars, they will be. It doesn't say blacks and Hispanics won't experience more force by police than will whites in similar non-lethal circumstances, they will be. But oddly, when more blacks are shot to death than are white suspects as a percentage of their respective populations, they are killed without prejudice (at least according to the police reports). So that's the thing about this that many might not have expected.
If I understand this right, it's not the over-abundance of killings which reveal discrimination, but it is the non-lethal actions which reveal discrimination.
So, while interesting to ponder, this does not close the conversation by any stretch of the imagination.