After a presidential election, some states (or local election boards) schedule their infrequent cleanup of voter rolls.
Typically they identify people who have not voted in 5 years (did not vote in 2012 primary or general and did not vote since) and move those from active to inactive status. Such a move has little significance on future vote counts.
They also check to state and federal death lists, which report about 98% of actual deaths and 0.5% of deaths to people still alive ... due to identity confusion mostly.
They also check for “movers”. Half the states allegedly participate in cross reference of new voters who are movers. Many movers in state. Surprisingly in state movers between election boards, or to addresses under the same election board are checked and “cleaned” less often than interstate.
Most vote fraud comes from voting in the name of a mover who is still on the rolls.
States “cleanup” results are usually available and need to be considered separate from results of those who literally change their party registration.
Of course, how many “movers” change their party registration at the same time? Typically, if one moves from a Democrat city to a Republican suburb the mover is more likely to change his party registration from Democrat to Republican for the sole reason that he is moving into a Republican area. Likewise when Republicans move into a Democrat city. Most people go with the majority and want to fit in.
Yes but the point is direction & trend. ALL favor GOP (even in declines) except CO, OR, and CA.