If I buy a research article, will I get any potential retraction statement or erratum free of charge? Or will I have to spend additional money to obtain access to it? I see that some retraction statements are behind paywall, so I wonder whether the people who bought the article also cannot access it free of charge.
I like the sentiment behind the question, but in my experience this depends entirely on the mechanics of the journal. If they replace the older version of the paper by a newer, corrected version -- and one may well think that they should never do this for anything other than publication errors on their part -- then your access should convert to the new paper. However if the corrigendum is a separate paper, then it's a separate paper: their online system doesn't know the difference and will charge you for it.
(It is not inevitably thus, of course, but journals would have to update the "sell pages" of their older papers to automatically point to the corrigendum, at least. I have not seen this. If anyone has, I'd be interested to know.)
One might regard this as the proof that buying journal papers individually out of your pocket is simply not a reasonable practice. I certainly recommend against it. If you need a paper and don't have institutional access to it: that's what friends / colleagues / collaborators / inter-library loans are for...