A common perception of the upcoming presidential election is that it will pivot on whether voters credit the recent uptick in employment as indicative of an economic turnaround, and thus support the President. Or whether they view record unemployment as a sign that a new approach to the economy is called for, even if the downward trend has been arrested.
This is consistent with academic models of presidential elections in which high unemployment at election-time, at least in the past, has predicted defeat for the party in power.