This is essentially the rationalization argument. It's not necessarily a bad one. But to put it in perspective, let's look at another area where racial profiling is said to occur in the U.S.: with blacks and crime.
In most large stores, the store employs undercover security guards who walk around to spot people who are shoplifting. Now imagine you had this position. The measure of your doing a good job is the number of people you catch who are stealing.
So, let's say you're watching the entrance to the store, and two people walk in and you want to follow one of them to see if they're stealing. One is a white woman, and the other is a black man. I bet nearly all security guards will follow the black man. Why? Because if you look at the statistics of who commits more crimes (rate, not number of people), it's black males who commit most of the crimes (I realize in my example, I confused the issue by comparing to a white female, but that's just to emphasize the disparity).
Now, you could say that because black males are always being watched and targeted, then that's why the arrest rate is so high (they're being watched and observed more). Maybe white males commit crimes at the same or higher rate, but just don't get caught because they aren't being observed as closely.
That's possible, but it doesn't apply when we look at the topic at hand: airline security (because you always know when the crime is committed because it results in an explosion/death/etc that's obvious). Let's look at all the airline-involved terrorist acts in the last two years, and there's a disproportionately high percentage of those commited by men of Middle Eastern descent.
So, if your goal as an airport security specialist is to stop as much terrorism as possible, then focusing on past data, you'll always target men of Middle Eastern descent.
This doesn't mean that people of other nationalities aren't capable of committing the same acts. But looking at past history, those of Middle Eastern descent do it at a disproportionately high rate.
I did a quick search on the web, and there's an interesting essay (with data) on racial profiling at: http://www.jeffsarchive.com/race/Race,%20Crime,%20and%20Violence.html