I would say to be cautious around correlation versus causation. I think when you look at traffic that's moved on our rail network across North America, it is overwhelmingly a function of macroeconomic conditions.
When you look at the period from 2014 to 2017, there was a period of relatively soft economic conditions followed by more robust growth throughout North America, especially in the U.S. Looking at that macro level volume, it's primarily a function of how the economy is doing.