It is widely acknowledged (in many funder mandates, for instance) that open access for peer-reviewed academic books in the humanities is a harder proposition. The labour invested in their production is quantitatively higher than for a similar journal article and degrees of cross-subsidy are often levelled across a Press's list in order to support scholarship that might not otherwise be economically viable.
Some of the economic solutions proposed to make monographs open access have included freemium models, print and electronic subsidy, institutional subsidy and consortial models (such as Knowledge Unlatched). In some of these modes it is suggested that publishers can recoup their costs through sales of the print version, or of specific other electronic versions of the open access book. So, for instance, I might give away the PDF (with open licensing even, perhaps) but sell a copy for Kindle. This is all well and good, but a slight problem looms on the horizon here that seems to need some clarification. The problem is Amazon.
Amazon claims to fight extremely hard for the consumer in lowering prices to all the items that it sells. While its practices are surely valued by customers, those whose revenue depends upon more than a pricing race to the bottom have expressed some concern here and this can be seen in Amazon's pricing policy for its Kindle titles. The guidelines for Kindle publishing state, with regard to list price, that "You must set your Digital Book's List Price (and change it from time-to-time if necessary) so that it is no higher than the list price in any sales channel for any digital or physical edition of the Digital Book". The question is: does giving academic work away in a different digital format count as a "sales channel"? If it does, then it is impossible to sell a Kindle version at any price above zero. I have heard publishers express this fear.
I would probably argue that a "sale" must involve some form of currency exchange and so, therefore, giving a version away would not count as a sales channel. Amazon seems to have disagreed in the past, though, and have price-matched to zero. There have been other claims, though, that they no longer undertake this practice. Publishers are worried about this. If they make a version OA, while relying on subsidy from print and electronic forms to cover their labour, I can understand that they would not want a zero-priced item appearing as an option just at the point when someone was going to buy the work (this is not the same as "hiding" the OA version and not making people aware that it exists).
This concern seems to be visible in the pilot collection of Knowledge Unlatched. From that list, these publishers do provide a Kindle version:
None of these, however, have been price-matched to zero, despite the availability of free copies. By contrast, though (and perhaps exhibiting this fear), these participants do not offer Kindle versions of their KU titles:
As far as I can see, all of these publishers except Brill have offered Kindle versions of other titles, so seem to have explicitly excluded Knowledge Unlatched titles from the Kindle offerings.
There is potentially a fear among publishers that intermediaries like Amazon will price-match to zero on OA titles, thereby removing digital edition subsidy as an option through a powerful sales channel. There is no evidence that this has happened so far with the Knowledge Unlatched collection. The KU pilot, however, is small-scale and simply may not be known to Amazon. In this case, it could be something that Amazon cottons on to later and could then force a zero-price match. Conversely, there is anecdotal information from self-publishers that Amazon is not zero-price matching at present. It also seems that Amazon's definition of a "sales channel" could be queried. I'm sure there's more to this and those with direct experience of Kindle publishing might be able to clarify matters. For now, though, this is what I could find.