"Not just because the police have a special inclination toward confabulation, but because, disturbingly, they have an incentive to lie." Actually there have been studies showing that police lie much more than the general public. E.g. here is a summary of one such study showing that: http://www.erowid.org/psychoactives/statistics/statistics_ar...
The only section of that page that might be relevant to what you just said was " police and military personnel, groups with more reason to deny past illegal drug use, were twice as likely to recant than the general population." You appear to be arguing that police are intrinsically more likely to lie, but this study doesn't support that, it actually supports the quote you are trying to refute.
I actually wasn't arguing that cops are more intrinsically likely to lie, only that they are more likely to lie. Although I do also believe that they are probably more intrinsically likely to lie. Why? I think you're right that of the non-police population, many would lie if they became cops due to the extrinsic incentives to lie. At the same time though, people who aren't comfortable lying about their current/past drug use are effectively banned from becoming cops. So while the study might not prove that cops are more intrinsically likely to lie per se, the fact that there is a significant disparity suggests to me that it's highly likely that part of that disparity is due to differences in who decides to be a cop. And, if it is the case that some significant portion of that disparity is due to selection factors, which seems likely, then my it would be accurate to say that they are intrinsically more likely to lie.