How could one argue that although racial profiling is almost always unfair, it can be effective, efficient, and necessary for our citizen's safety?
This argument is one that makes factual claims, and so to argue it one needs to do research concerning the racial makeup of the largest portion of terrorists who have actually attacked the United States. Although the single most dramatic and memorable attack was 11 September, it was unusual in being carried out by Arabic terrorists. The great majority of terrorist attacks in the United States have been planned or carried out by white males. Thus there is some justification in devoted extra security to searching white males. However, since certain attacks have been carried out by non-whites and females, this might not be an effective strategy. It might just lead to more gender equity in terrorist organizations (I'm not sure whether we should consider gender equity among white supremacists, e.g., a social good or not). Your most important task would be some statistical evidence showing a drop in the number of attacks in jurisdictions that use racial profiling compared to that which do not.