There are two levels at which this can be answered, I suppose: what were the personal motivations for offering/agreeing a draw, and what were the objective features of the position that grounded the decision? I think you're asking mostly after the second, but I've read many comments lately voicing frustration with agreed draws in this match, so I hope you don't mind me giving an answer that speaks to the first level as well.
First off, the position after 21 moves was as follows:
[fen "r2r4/4bkpp/5p2/p1Bbp3/7P/1P1P1N2/P4PP1/R1R3K1 w - - 0 1"]
1. Bxe7 1/2-1/2
Here Anand played
22. Bxe7 and offered a draw, despite being a pawn up in material and having a significant time advantage (with still 18 moves left to the time control after what would have been Gelfand's obligatory recapture
In the post-game press conference, Anand was asked if he considered trying to play on his time advantage at the end, using it to hopefully induce errors on Gelfand's part. His reply:
Bxe7 Kxe7, black's play is very easy. I didn't see any point in dragging it on from there."
In reply to a later question, Anand elaborated further on this and other agreed draws from the match:
"We only draw when it's obvious, at least to us, that the game is going nowhere."
One journalist spoke up for the amateur spectators, in the spirit of your question, asking why the contestants couldn't play out some more moves for the benefit of those who don't see why this position is a draw. Gelfand fielded that question, answering:
"First and foremost, here, we are competing in the match for the crown. We are not here purely to entertain spectators. Second thing, there are about a dozen professional, highly-qualified commentators, and with a little bit of effort they could explain to the thousands of spectators elsewhere the nuances of the game."
I think this is a fair point; his and Anand's job right now is to make chess decisions based on personal competitive interests, and not to explain the games either verbally afterward or by playing on in a position each recognizes as clearly equal, and offering no realistic prospects of an advantage for his side.
Which brings me to the second level of your question. I am not a highly-qualified professional, of course, but I think the concrete factors leading to Anand's draw offer are that, though he is a pawn up, it is a weak isolated pawn on d3, one which is hard to make use of, or to be rid of by advancing to d4. And his 2-to-1 pawn majority on the queenside is very immobile, with Gelfand's bishop and rooks having plenty of space and open lines to operate with and restrict Anand's pawns. Furthermore, Gelfand has not a single structural weakness in his position. In a nutshell, this is why Anand's extra pawn doesn't mean very much, and why Gelfand's play is easy enough in this position that Anand didn't consider the time factor to be terribly important.