I found a link supporting this, but I wonder if there is some cherry-picking going on. Sure, right after the nominee changes from an unpopular one to a popular one, you are going to see some registration by people who like the new nominee. Also, some of that link is about women registering more than men, but what matters is their party, not their gender or race. To (yes, I know this next phrase is rightly mocked in some contexts) do my own research, I looked here:
Pennsylvania Party Registration Totals and Changes, Jan. Through Sept. 2024
The party changes look quite bad for Democrats – both year to date, and last week alone. For the week ending September 30, 2024, 2,803 voters changed from Democratic to Republican, and only.1,301 went from Republican to Democratic. Year to date, 54,308 went from Democratic to Republican, and only 21,995 went from Republican to Democratic.
What about new registrations? I do not see that by itself, but total Pennsylvania registered voter numbers, as of June 24, 2024, (before nominee switch) can be seen here:
Internet Archive Backup.
Using my last two links, I find:
June 24, 2024
Democratic registered voters - 3,893,339
Republican registered voters - 3,517,502
Democratic registration advantage - 375,837
September 30, 2024
Democratic registered voters - 3,941,347
Republican registered voters - 3,608,032
Democratic registration advantage - 333,315
So – not a total disaster for Democrats, but Pennsylvania data suggests that any Harris enthusiasm registration bump was brief and relatively minimal.
Other states? Not sure. I can find newspaper articles proving whatever I want 