This has been fairly common in my experience under certain conditions (none of which you should take personally), but here are some justifications:
1. It's a buyer's market. Given other fairly similar qualifications, the ABD is going to lose out to the PhD (most likely in the post-doc position). This is more likely to happen as the number of applications increases and the quality at the top of the pool increases. But with smaller pools where most applicants have ABDs (AND are the most research productive), it would be silly and counterproductive to implement such a policy. Bottom line: Institutions will only use this if they can afford to be picky.
2. Building from number 1, post-docs are more likely to have pubs that show research independence. Again, don't take this personally and as law, but IN GENERAL candidates with PhDs can show they are not just doing their advisor's work and have more first authorships. Remember, it's not about the number of pubs, but what you did on them and whether that research is truly "yours" or not.
3. Building from number 2, ABDs are a productivity gamble. Sure, the department can take your word that you are almost finished, but what if life circumstances occur and you don't finish on time? There are departments I know who would never consider ABD candidates because this happened before. Even when if someone finishes, but after a year or two on the job, it bodes very unfavorably on the department. The new assistant professors spend too much of their research time finishing their dissertation work instead of building a new lab at the institution.
4. ABDs are also a 'professionalism' gamble. Sometimes, compared to older and more experienced counterparts (and hey, this can be even a year), ABDs are a bit emotionally and professionally immature; they are not sure how to set up professional boundaries between themselves and the graduate students (e.g. they have trouble 'identifying' with faculty because they don't truly feel like one yet) and have trouble adjusting from the grad student life so quickly. Again, individual variation and all that, but this happens much more frequently when you hire an ABD than a post-doc or someone who has been out of the grad life longer. If a department has experienced that issue before (and I know another where it has), they may be more inclined to implement a 'no ABD' cut on future searches.
5. Given all the above, if you have a large applicant pool you need to cut down, and it is filled with plenty of exceptional PhDs in-hand, I can't see any reason why to consider ABDs. Technically, they have one less (extremely important) qualification than the others, and if the department has been burned in the past, they may decide to use this as a heuristic to prevent hiring disasters. Of course this isn't fair based on specific individuals (e.g. some ABDs are far more ready for positions than some post-docs), but you can see how it can be justified.
To provide a little perspective, some other fields (like the natural sciences) won't even consider candidates who haven't had at least two post-doc positions….that's a bare-minimum of 2 years past the PhD (but often more, given positions are typically 2-3 years)! My significant other is in such a field, and his colleagues are astounded that candidates in social sciences are even interviewed ABD. They really can't even wrap their heads around it. So, I suppose the general point is to view the glass as half full; that many institutions DO still consider ABDs! That may change in our field too over the next decade or so….