Because we still don't know exactly how the brain works. That's why we do research. If these authors knew the answer ahead of time, neither of them would have done the study.
These are research papers, they are showing findings that are consistent with different models. Maybe the real model is somewhere in between, maybe there are context-dependencies, maybe one measure was more sensitive than the other for a particular result, maybe statistical thresholds caused one paper to emphasize one circuit and another to emphasize another.
There are other mediating effects like arousal and salience that can bias results; responses may not be linear so the degree of effect may be important.
Brain regions can excite or inhibit other regions and those connections can be reciprocal, so it is often not clear if decreases in one area are causing or caused by increases in another.
There are more recent papers like Bartra et al 2013 (citation below) that have tried to compare these different studies, so that's a good place to start for further reading, but you will find that they still do not have a clear answer.
Bartra, O., McGuire, J. T., & Kable, J. W. (2013). The valuation system: a coordinate-based meta-analysis of BOLD fMRI experiments examining neural correlates of subjective value. Neuroimage, 76, 412-427.