I have some descriptive, quantitative findings that I think are notable even absent any multivariable regression modeling. The findings contradict some received wisdom in my subfield, and I think other scholars' rush to use advanced methods lead them to miss these simple aggregate trends. What are some journals you know of that publish descriptive results? Race/ethnicity, criminology, education, or general soc journals would fit, topically. Thanks!
What journal might publish descriptive findings without regressions?

I think it depends. If the descriptive findings rise to the level of what is called a causal process observation in the process tracing literature and you can make a case study type thing out of it then it's a qual article in a sense. If it's a single finding I think you could do a research report but it willl be a tough sell. It's conceptual and conceptual stuff is much harder to place. What about Contexts?

Fixation to that a paper must have a regression modelling, is indeed annoying. Oftentimes, the essential descriptive pattern is much better interpretable, and with less risk of bias, than ritual of regression for regression's sake. And regression is indeed advanced statistics for a layperson. We should remember this especially if we aim for relevance outside of our small circles. In addition, I am not too optimistic for sociologists either... Often you can communicate much clearer with more descriptive strategies.
To answer the original question. Demography (as a field) has traditionally appreciated descriptive results. Of course, your description does not necessary meet the substantive criteria there. Demography (the journal), on the other hand, publishes quite wide variety of quantitative sociology, so it could be an option. However, your descriptive findings should better be interesting to aim there!

This discussion is depressing. Let’s say the dscriptive data calls into question a widelyheld belief and the data is contextualized by: 1) reviewing just how common this belief/assumption is and 2) drawing out the implications of this belief potentially being false. This is what I assume the paper looks like. If so, that’s far more important and interesting than your average regression that confirms priors or shows the obvious.
If done well, why is that not publishable as a research article? Why would s/he have to attach a regression to it? I’m not saying this advise is wrong, I’m saying it’s absurd that this advise has to be given. How silly this field has become.