NJ 2020 results versus 2016

[jman123]

I have been studying the Atlas NJ town by town election results and swings and i have been struck by certain things overall.
First, Trump gained in raw percentage points 0.25 more over 2016. The vast majority of 2016 third party vote went to Biden so he got 57 percent compared to clinton 54. Trump got 41 percent in 2016 and 41.25 in 2020 in NJ. One is struck by the large swing in highly educated white suburbs to Biden. This alone would have lowered Trumps total percentages to below 40 percent but something occured that i never expected. If you look at the town swing map, the large Republican swings occured in minority heavy, predominately hispanic , inner city urban areas. 10 -25 point swings to Trump. This counterbalanced somehow trump's slide in suburbs to maintain trump above 40 percent statewide.

What led to his improvements in urban areas?
Inroads among hispanics?  What may have led to this? In 2016 Trump posted poor numbers in NJ urban areas but held his own in suburbs. Why this reversal?

[Pericles]

Trump made gains with Hispanics nationwide, so it's not shocking that this also happened in New Jersey. It still isn't clear why he got such strong swings with Hispanics-some theories are that voters are now polarising along ideological and educational lines instead of racial ones (Trump does seem to have won minority voters who already had conservative views), an anti-lockdown backlash (why this would cause such a difference in swings based on race is unclear though), Hispanics always swing to the incumbent (true but it doesn't explain the extent of the swing), and Trump's overall image being appealing without him focusing as much on immigration that turned these voters off (maybe). There are probably some other theories and more detailed and nuanced explanations. One thing worth noting is that while this swing is still weird, it was predictable from looking at the polling crosstabs in 2020 that showed Trump making big gains with Hispanics. Unfortunately looking at the crosstab averages isn't foolproof, we were supposed to have big swings among seniors and non college-educated whites to Biden (the latter was polling to the left of both the 2016 polls and results), but that didn't happen.