I agree (hence my original statement: "With such a small percentage of the workforce unionized, it's highly unlikely those employees enjoyed any additional bargaining power or job security that could result from having a significant fraction of the workforce unionized. With such a small fraction I don't think you can draw conclusions in either direction.")

However I then, uh, went on to draw a conclusion anyway in contradiction of my own statement saying we shouldn't draw conclusions because of the sample size. Methinks it's time for more coffee :).