Much ink has been spilt on the need to encourage projects that reduce net carbon emissions (with money) without paying people to do what they were already going to do.
The trouble with the requirement that people be paid for "additionality" is that it's too hard to separate real from imagined moves that improve the carbon balance.
This accounting and fraud problem is slowing or blocking a lot of projects, because buyers of carbon credits do not want to pay for projects that may not be certified as "additional."
This is stupid.
First, projects are not getting done (or, even worse, are getting credits for NOT happening), which means that the desired impact -- less carbon -- is even farther away.
Second, it's not morally defensible to exclude projects "that would have happened anyway" from receiving payment. If someone has a forest, why not pay them for not cutting it? Even if they were not going to cut it anyway, it seems nice to pay them for having such a useful asset.
The only reason that I can see for keeping additionality is that some bureaucrats or enviros want to maximize bang for their buck, to take the "free" stuff for granted and pay for additional stuff. This tightwad perspective sounds good on paper, but I don't think it's good, for the reasons I mentioned above. Further, I can tell you that the price of keeping something that you want to keep anyway is pretty low. With competitive markets and bidding for carbon sinks, it's a sure thing that the cost of carbon deterrence will be low.
Bottom Line: Pay people for doing things we like, even if they already want to do those things.
On a related note, read this excellent article on the value of ecosystem services, why we need to account for them and how much we should pay for them.