In late July, Hunter Harrison, the CEO of CSX railway corporation, declared that his railroad, founded to haul coal, and still getting a fifth of its revenue from coal, would no longer invest capital in its coal business – neither new locomotives or track investments designed to support its coal business. Harrison was blunt as to the reason. “Fossil fuels are dead,” the CSX CEO said. “That’s a long-term view. It’s not going to happen overnight. It’s not going to be in two or three years. But it’s going away, in my view.”

The same week, Morgan Stanley reported that Trump or no Trump, “we expect the US to exceed the Paris commitment of a 26 to 28 percent reduction in its 2005-level carbon emissions by 2020.” The investment banker cited as the reason its conviction that with a few exceptions, renewable energy would be cheaper than fossil power by 2020 in every important market.

But such radical market share growth for non fossil fuel energy is not reflected in most top down projections. DOE, for example, continues to project very slow declines in US climate emissions over the next decade. More alarmingly, an analysis derived from BP’s most recent “Statistical Review of World Energy” looked at the percentage of total world energy derived from fossil fuels and at the rate of growth of such energy use, and concluded that there was no evidence whatsoever of a turning point having been reached.

Why such sharp divergences of opinion? Is one side cooking the books? Probably not. There is a well known difference in economics between leading and trailing indicators. After all, this year’s consumption of gasoline, or combustion of coal, is largely driven by investments made in cars or power plants a decade or three ago. But this year’s investments in various kinds of vehicles or power generation will cast an emissions shadow for years to come.

Actual consumption (or emissions) is a trailing indicator. The decision by CSX to no longer to invest in its coal business won’t drive coal demand down next year at all – but over time, if sustained, less coal will come to market – it’s a leading indicator. The problem with leading indicators is that no one is a sure-fire harbinger. Sale of electric vehicles is one leading indicator in the transportation sector. But so is the percentage of US vehicles sales that are SUV’s. Both might go up in a given year.