Then, when you realize that toll roads are correlated with population density, perhaps you could realize that toll roads are then only also correlated with politics of state, not caused by them.

For example: If liberals loved toll roads, then California, Oregon and Washington would be full of them. But only the NorthEast is, right?

So your hypothesis falls apart here anyway.