For a direct reference on the contribution of these two forces see here.
If it helps, think of the double helix of dsDNA as a more rigid structure than ssDNA, which flops around and generally is capable of a lot of steric freedom that dsDNA doesn't have. This freedom means that the bases of the ssDNA backbone may not always be in the correct orientation for stacking, because the two strands of the helix aren't forcing the bases together into that orientation. "Usually the DNA double helix is considered to be a relatively rigid structure. Certainly it is much more rigid than its component single strands." (see the review for a much more detailed discussion of this issue)
It's true that base stacking still occurs in ssDNA, it's just not enforced in the same way. It also appears to be somewhat sequence-specific, occurring more in polypurines than polypyrimidines, possibly because purines are bulkier molecules with more steric hindrance.
For a related conversation see here.