Some journals have a very smart strategy to ensure that reviewers reject as many articles as possible. Reviewers are told that if they recommend an article for publication, they have to proofread, make sure the bibliography is in the right format, correct spelling and stylistic errors, and make the text ready for publication. Since most people write extremely carelessly, this is a mountain of work. Nobody wants to do it, so it’s easier to reject.
In short, it takes about 2,000% more time to accept than to reject.