The GOP jump seems notable and steep, but the Democratic dip feels like it’s in line with the normal noise and variance that you get with a reoccurring survey.

https://news.gallup.com/poll/507230/fewer-say-sex-relations-morally-acceptable.aspx

You’re going to get normal peaks and valleys with a reoccurring survey. The respondent pools are never identical year over year. And with progressives, who skew younger and don’t answer phone surveys, you’re going to have even more noise.