Stephanie, the three answers so far all raise valid possibilities. I've personally seen evidence of lazy census takers, where all locations were local, even for confirmed immigrants. And ages can vary widely from one census to the next -- I found one case where a person had only aged 3 years in ten -- and at the next census had aged another 16 years in ten! I think this could also be attributed to who the census taker actually talked to. If the interview was done with an older child, for example, while the parents weren't home... or perhaps a parent-in-law...
With names, there are several possibilities. I would be willing to bet that most census takers, especially in the early 19th century, rarely asked for someone to spell the name for them -- assuming that the responder didn't know how to spell it themselves, especially in rural areas.
There is also the added complication of the person who transcribed the record for the index. I've found numerous examples where the transcription was wrong, not the actual record. This obviously applies to any handwritten record, not just the censuses.
And the shifting boundaries can also cause real headaches. I have a case in my own family where some migrated to Missouri. The records for that family shifted from Missouri to Iowa and back, even though they never actually moved -- just the boundary moved!
I have a suggestion -- make a list of all your current sources, with derived date of birth and place. Then try to determine how accurate each of those sources might be -- how consistent are they? How close in time are they to the actual event? If the record is only one of several on a page, does anything else on the page suggest that it is more or less accurate?
And if you haven't already, try finding the records for his siblings. Often, that will help locate the family as a whole.
