If it's a paper from an RA job with a professor then I'd recommend no, even if the job posting allows it (which they often don't). A co-authored RA paper can only serve as a good signal of your competence at RA data cleaning or lit review, but that information should already be included in the LOR from the professor. On aspects like communication ability, research intuition, or econometric proficiency, it's an ambiguous signal when the quality is good because your advisor could have guided you, but a clear negative signal if there are errors/mistakes, because neither you nor your advisor noticed them.
What typical empirical RA positions look for are sufficient programming knowledge and English literacy to get started on RA work immediately, not to gauge the quality of your best research previously produced (that's a job for journal reviewers or PhD adcoms). It's possible that a mediocre/incomplete solo authored paper can signal enough about your knowledge to surpass a supervisor's standard for RA hiring, while a much better co-authored paper can't provide the same reassurance.