Generally speaking, good sociologists understand that sociological concepts are abstract approximations, not definitive absolutes. Unfortunately, in popular discussions, including sometimes political discussions, these concepts are frequently used in improper manners.
The best example of this is, perhaps, privilege. There are a few ways in which this concept is misunderstood or misused. The first is that privilege is a sort of an average. When speaking of a group’s privilege, it is often forgotten that each group carries certain general advantages and disadvantages, and it is disingenuous to pretend that we can simply average these out. Often, the advantages and disadvantages are highly situational. There are ways in which being a woman is disadvantageous, for example, but there are situations in which it is an advantage: women are not (or at least were not historically) conscripted in most countries where men are, for example.
Another way in which privilege is misunderstood is that many forms of privilege will only benefit a certain subset of the group. Often, these advantages may be quite irrelevant. Take the racially biased stop-and-frisk practice in NYC. This practice disproportionally targeted non-whites, hence, this became a part of white privilege. On the other hand, this privilege did not benefit a white person unless they happened to be in NYC. In addition, while there was racism involved in the application overall, it would not be proper to assume that any given case of stop-and-frisk was racially motivated. Some of them could have been legitimately done at random, or done because there was some cause for suspicion.
By a similar token, when talking about people who die at the hands of the police, this disproportionately happens to black people (though it does also happen to white people). However, we cannot conclude from this that a given instance where a black person dies during an arrest attempt is the result of racism. More information is needed before any kind of conclusion like this could be made.
Additionally, it is often forgotten, or overlooked, that privilege refers to how easy or hard a time groups have it on average, which says next to nothing about a specific person. That ‘white privilege’ is higher in America than ‘black privilege’ is probably quite accurate, but this says nothing about how easy or difficult a time a given white person will have compared to a given black person. Black people often have a harder time with police than white people. However, a white person living in a town with a black sheriff who is bigoted against white people might have a much harder time dealing with the authorities than the black people in that town.
Another problem is that ‘privilege’ is often not a good word for what is going on. Let us take the male/female wage gap. This gap is largely explained by the types of jobs and experience in which males vs. females are employed. Now, the question becomes, why are they in these jobs? Is there really some pro-male bias going on in employment, or is this the natural result of the choices which males and females make, on average? In other words, is there an actual advantage to being male, or do males just tend to prioritize higher pay more than women (we know the latter is the case from studies). If the latter explains the entire difference in employment, then there’s no widespread bias, and hence ‘privilege’ is a rather silly word for describing the situation. Of course, even if there is no widespread bias, there can certainly be some individual employers that are biased (both towards women and men; consider that the Huffington Post posted a picture of their editorial room, bragging about how it had no men in it – it also had only white women). But is ‘privilege’ really an appropriate term if the gap is explained entirely by voluntary choices?
Now, related to the above points about how certain types of privilege often fail to affect more than a subset of the group, it is frequently the case that discussions of ‘groups’ include groups that are far too large, in the sense that many sociological aspects of the group really only apply to certain subsets. This is the case not merely with regard to privilege, but any sociological concept. ‘White American’ is an incredibly broad group. The cultures and experiences of most white software engineers in Portland are not very comparable to a poor white Appalachian coal miner, or a Texan farmer. ‘Black America’ is likewise extremely broad. So, for many sociological applications, ‘black’ and ‘white’ are not necessarily very good divisions.
To put all this into perspective, we can make an analogy to tracking the weather in a given city. We could ask, what is the average temperature during the month of June? Perhaps it comes out to be 25 degrees Celsius. However, it is likely to be radically different from that on many days: it might be 15 degrees one day, 30 another. Trying to apply broad sociological concepts to make statements about individual instances (e.g. claiming every black person who dies during an arrest attempt died because they were black) would be like claiming that the temperature must have always equaled the average for the entire month. In addition, when asking questions of average temperature, some numbers are more meaningful than others. Asking what the average temperature was over the entire year, for example, does not provide us with much useful information if we are trying to figure out what the weather is like at a particular time of year. Similarly, for many questions, in sociology, it is important to break up the groups by divisions which are meaningful for answering the questions which are under current investigation.
Good sociologists understand this, of course. They understand the limitations of sociology, and how to properly interpret data. But popular debate needs to become more educated on this topic.