So far, Trump’s job approval is off its peak, but it is still at a level that would have been considered a peak a few months ago. We’ll see how this holds up over the next few months.

The economy, however, is another matter. Trump’s best argument for reelection was the strong economy, and the February jobs numbers suggested that we had shaken off the doldrums of last fall. But with large portions of the world clamping down on travel and businesses being uncertain about the future, a drop-off in growth seems unavoidable.

If this virus follows a best-case path, we might see a sharp rebound in the third quarter – we may recall this from the debate over the “V-shaped” vs. “U-shaped” recovery in 2009. The problem for Trump is that political science models have generally looked to the second quarter as the key marker for explaining presidential elections. There isn’t unanimity on this, but the basic idea is that economic conditions take a while to soak in.

In other words, if things are getting better in September and October, people won’t perceive it until after the election. This can explain why President George H.W. Bush didn’t benefit from the improving economy in the second half of 1992, and why John McCain only lost by seven percentage points amid the 2008 financial collapse.