First of all, a decision on whether a corrigendum is issued is up to the publisher, who usually acts according to the recommendations of the journal editor.
Consider Elsevier's "Policy and Best Practices: errata and corrigendum" (emphasis is mine):
A corrigendum refers to a change to their article that the author wishes to publish at any time after acceptance. Authors should contact the editor of the journal, who will determine the impact of the change and decide on the appropriate course of action. Elsevier will only instigate a corrigendum to a published article after receiving approval and instructions from the editor.
A more detailed process is described for JMIR publications, where they describe potential outcomes:
Requests like "I forgot to acknowledge somebody" or "there is an error in one of the authors' names" are what we call a "Discretionary Correction", as it is an oversight that is the authors' responsibility but not severe enough to affect the validity of the paper. We can correct it, but reserve the right to charge a fee of $190 for publishing the corrigendum, correcting the original article and linking it to the corrigendum, and resubmitting the article to various databases and/or making changes to the PubMed record.
We do not think the error you have mentioned is a problem and requires any action/correction.
The error is a minor layout change and can be made without publishing a corrigendum.
So, for JMIR, the type of error can lead to different results: publication of corrigendum (with fees paid by the publisher or by paper authors), rejection of the corrigendum inquiry, or in-place edit of the publication without formally publishing a corrigendum.
Therefore, the process is definitely field-, journal- and, probably, even editor-dependent.
In your case, the error in the in-text citation is very unlikely to be considered severe, especially, since the citation in parenthesis is correct.