Did the conventions give either candidate in the presidential election a lift? Emerson says yes, and might have even done the all-but-impossible — give Donald Trump a positive favorability rating, at least briefly. Emerson’s findings appear to correlate to trends in other polls, even if those don’t hit the same amplitudes as Emerson finds.

The head-to-head bounce against Joe Biden doesn’t really amount to much. Emerson reports that Trump narrowed a four-point Biden lead to two points, essentially a margin-of-error difference:

Since the Emerson College July national poll, President Donald Trump has tightened the presidential race to a two-point margin, and is now trailing former Vice President Joe Biden 49% to 47%. In July, former V.P. Joe Biden held a four-point lead over President Trump –– 50% to 46%.

Spencer Kimball, Director of Emerson College Polling explains “the Republican convention gave Trump his most positive week of news coverage which likely attributes to his bounce in this month’s poll and increasing job approval.”

The job approval move seems more significant, but the demos are worrisome:

For the first time since he’s taken office, Trump’s job approval rating is approaching a majority, at 49% approval and 47% disapproval. This is a jump of four points since July, where Trump had 45% approval and 51% disapproval. Trump’s approval is the highest among those aged 30-49 (55% approval), and lowest among those aged 18-29 (37% approval). His approval among those aged 65 and older has fallen almost ten percent since the start of the COVID-19 pandemic, from 50% approval in February to 41% approval today.

Trump really needs older voters to turn out massively for his re-election in order to compete in November. That was one of his stronger demos four years ago, and the decline in approval there should alarm the campaign. Emerson’s otherwise cheery-ish poll shows Biden leading  by fourteen points among seniors in the head-to-head match, and their voting reliability makes that demo key to electoral success.

On the other hand, there are a few demo highlights for Trump, too. Biden only gets a 77/19 split among African-Americans; if Trump and the GOP score 19% of that demo, it would spell disaster for Democrats. Trump also does slightly better within his own party (83/14) than Biden does in his (79/18), which might indicate an enthusiasm problem for Biden. Bear in mind, though, that demos and subdemos have a higher MoE than for the overall poll, which clocks in at ±2.4 points.

This is one poll, of course. How does it fit in with the trends? Again, it does tend to mirror what RCP’s aggregates show in the election race, especially over the last month. Biden hasn’t really lost ground, but Trump picked up support, especially over the last couple of weeks — after bottoming out in July. Meanwhile, Biden’s been static since clinching the nomination, much as he was static in the Democratic primaries:

On job approval, the trend is there too, but the news isn’t nearly as cheery as Emerson sees:

There is one last point to keep in mind, too. These are all national polls, not battleground state polls that matter most in these elections. And in 2020, even battleground state polls will struggle to come up with reliable likely-voter models that will offer a real predictive look at the election. If Trump is as close as Emerson posits in a national poll, one can reliably assume that Trump’s strength is likely growing in those battleground states to an extent that the big blue states of California and New York no longer swamp out Trump’s level of polling support.