As mcmurtry pointed out,because more people VOTED., The reason Trump won the EC in 2016 was because Dems (in MI, PA, and WI) stayed home or voted 3rd Party. A lot even voted for Trump...
But after 2 yrs in office, all 3 of those states were done with Trump. So in the 2018 midterms, all 3 of those states threw out GOP Govs and failed to elect any candidate Trump endorsed in a contested race...
And it was easy to see why, if you knew where to look. In Jan 2017 when Trump was inaugurated Morning Consult started a daily tracking poll of Trump's popularity state by state. They compiled monthly numbers based on their monthly survey of the same 5000 voters pool each month to track any fluctuations...
So in Jan 2017 Trump's approval rating (approval vs disapproval) in MI was 48%-40 (+8), in PA it was 49%-39% (+10), and in WI it was 47-41 (+6). This was the honeymoon period when the polling suggested these voters were generally more approving than disapproving of Trump in the states where he won the 2016 electoral vote...
But by the 2018 midterms, the drop in Trump's popularity among the same core group of voters in those 3 key states was incredible. He was now underwater in all 3 states with MI being 43-52 (-9) overall -17% decrease, PA being 46-51 (-5) overall -15% decrease and WI being 43-53 (-10) which represents a -16% drop in Trump's overall popularity from Jan 2017 to the Nov 2018 midterms...
Before you mischaracterize that as meaningless polling, you need to realize that election results in Nov 2018 in all 3 of those states reflected that decrease in popularity. The Dems didn't even have viable candidates for Gov in place on election night 2016 in any of those states, and by 2 yrs later the Dems had won every single Senate race and the Gov, as well as House seats in contested districts in EACH of those 3 states...
The reason all of that is significant is that those 3 states are where Biden WON the 2020 election. Your sophomoric claims about Trump's "popularity" aside, clearly Dem voters were much more motivated in those states in 2018 and 2020 than they were in 2016. You live in PA, are you unaware that a huge reason Trump lost PA is that he lost significant margins in red districts he won from 2016 to 2020?
Look at Butler Co, which Trump won in both 2016 and 2020. It's a suburb of Pittsburgh, or at least adjacent to Allegheny County. In 2016 Trump won the vote there by more than 37 % (66.7% to 29.2%) This is a GOP county, and you only believe Dems "cheat", which is why I'm showing you this county. Close to 100,000 votes cast with Trump getting 61,388 and HRC getting less than 27,000 around 26,800 or so.Three other candidates dividing 4000 or so votes...
In 2016 those types of rural margins helped Trump to counter Clinton's lead in Blue areas like Philly and Pitt. But in 2020 while more people overall voted in Butler (close to 112,000) Trump's margin went from nearly 38% to 32.5 %. While Trump got 74,300 votes Biden got 37,500 and that same loss of victory % was reflected in many other red counties in PA. Which meant far fewer votes from rural areas for Trump to try and hold off huge Dem amounts in Phil and Pitt.
That's how Trump's 1.2 win in 2016 turns into a roughly 1.2% Biden win in 2020. It's also important to note that with 5 candidates receiving votes in 2016 Trump was only at 48.8 % of the total vote share. But with only a two-man race in 2020 Biden was right at 50%. Again all of these stats point directly to less % of votes FOR Trump among PA Trump voters in 2020 than in 2016...This further validates the MC polling numbers...
As to your rally nonsense, most Biden voters voted EARLY (in person or mail), which means while Trump was holding rallies most Biden voters had already voted. Biden voters weren't concerned with "rallies"- I've never been to a political rally.
Equally telling is that Biden voter's enthusiasm was reflected in huge primary numbers over Clinton in 2016. In primary after primary total Dem vote numbers in 2020 exceeded the numbers that had voted in 2016. In a state like MI where Sanders won in 2016, Biden's numbers in 2020 eclipsed the total votes cast for Clinton/Bernie combined in 2016. And both Clinton and Bernie held rallies equal to Trump when they campaigned in 2016...
But the biggest fallacy on your Trump rally theory is the Trump rally themselves. Significant people at Trump rallies were "groupies" who attended multiple rallies. And MANY of them were not locals and didn't reflect the shape of the actual election in the area the rally was held.
An example was a HUGE Trump rally held in Omaha, which had Nebraskans as well as people from surrounding states like Kansas, Iowa, Ok, etc... But Neb breaks down EVs by congressional districts, and Biden beat Trump for the EV that the district Omaha includes awarded.
In other words, Trump held a rally in Omaha, but very few of the people attending were local to Omaha, and in general, they came from deeply red states, to begin with... That's another reason you have to take your "rally size" theory with a grain of salt...People at Trump rallies voted for Trump, but it's where they actually voted that matters...
morningconsult.com