Journals should avoid plagiarism
A spectrum of plagiarism (Turnitin) ranges from cloning someone else's entire paper to sampling some pieces with a few word alterations.
There are times when an author may want to build on his or her previous work. However, many authors, and even editors, may be unaware that doing so improperly may actually be self-plagiarism and as such runs afoul of scholarly ethics.
Self-plagiarism can occurs when an author submits several publications based on a single study, submits the same work to multiple publications without disclosing the fact of multiple submissions, or uses text or ideas from past works.
The type of self-plagiarism most likely to occur without a plagiarism mens rea is probably text recycling, simply because using one's own words or ideas is not intuitively a violation of plagiarism ethics. After all, how does one steal from one's self?
It is true that text recycling is not plagiarism in the usual sense. Rather, it may be a violation of trust the publishers and readers have that what they are reading is original scholarship.
This is not to say that an author can never reuse previous work. Here is an example of a 97-page article that was later pared down to 10 pages for a second publication:
Image from Google Scholar at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2799334.
(Note that best practice would likely be to use an alternate title for the second publication.)
Reusing work in this way can make it much more accessible to readers, and can allow the author to provide a more focused look at the key ideas.
It is also possible that the author wishes to further develop the ideas of the original article and writes a second, more in depth article. Here is an example where an author further developed and published a paper that he initially wrote as a master's student:
In the published version, the author's note on the first page states that the article is "a revised, updated, and expanded version" of the author's earlier paper.
The key to avoid an ethical violation and upsetting your editors and readers is to adequately disclose any work you are reusing, even if it is your own. If in doubt, ask your editors for guidance. It is possible that they have not dealt with the issue either, and you may need to work together to develop solutions on the road to publication.
See Dealing with Self-Plagiarism (a.k.a. Text Recycling) in Law Reviews, by Yasmin Sokkar Harker, Student Liaison Librarian, CUNY School of Law & Benjamin J. Keele, Research and Instructional Services Librarian, Indiana University Robert H. McKinney School of Law, for an excellent review of the issue.
In addition, this article on Self-Plagiarism by law professor Josh Blackmun discusses his own experience with issues of self-plagiarism in the law review context.
If a prior work has been published, another consideration is whether the author or the publisher holds the copyright to the prior work.
Online services don't cover everything.
The commercial plagiarism-checking services can save you a lot of work, but nothing is perfect. Bear in mind that an author could have plagiarized from a book or article that's not in the service's database. If you suspect plagiarism, check further, even if the program didn't find a problem.
Watch for dates.
If an author submits a paper that was supposedly just written, watch for sections that cite old sources (e.g., C.F.R. sections or U.S.C.A. pocket parts from a few years ago). That could be a sign that the author "borrowed" someone else's footnotes.