You've asked Unrepresented to prove something that is essentially unprovable. Not because the one variable did not cause the other, but because causation simply does not exist. If we could manipulate one variable while leaving all others constant we could prove that a rise in one variable seems to occur simultaneously with a reduction in the other variable. That is as close as we can get to establishing causation which, as I say, in an objective sense does not really exist. When a cue ball strikes a numbered ball and the numbered ball starts moving, we think the cue ball caused the motion. But really, all we have is two events that seem to regularly, reliably, occur one after the other. We still have not found causation. What we have found is convention.
In any case, the real wages (adjusted for inflation) of hourly employees have been falling for many decades. No credible economist would suggest otherwise. In that same period, union affiliation has fallen. These variables are, without a doubt, strongly correlated. I would not expect a rise in wages to cause a rise in union affiliation. However, a reasonable person might expect a rise in affiliation to be correlated with a rise in wages. That connection is certainly not counter-intuitive.
Given the brute fact that real wages and union affiliation are strongly correlated, I wonder if you can offer a reason we should dismiss the theory that affiliation affects wages. Particularly in light of data which regularly shows that union workers wages are almost invariably higher than those of non-union workers who perform the same tasks.
