I think there are two responses to it.
First of all, the action plan must come with a plan to also collect disaggregated data because you want to make sure that whatever action you come up with will benefit different communities and not just benefit some communities. The only way that you will know is if you collect data that is disaggregated, looking at different communities.
The second part is what we call targeted universality. You may have a general action plan, but it's identifying certain communities that are most at risk, with special attention being put to those communities within the overall plan. An example of that would be if you have an action plan looking at different components, one component might be on the correctional system. We know from the data that the communities that are most affected would be indigenous and African Canadian. You have an overall plan. You may have something that says something about the correctional system, but within that component you may also have certain targeted measures for those communities.