Great idea from the Association of American Universities (AAU) and the Association of Research Libraries (ARL) for an institutional subsidy for publication of a scholar’s first book: http://www.arl.org/publications-resources/3280-aau-arl-prospectus-for-an-institutionally-funded-first-book-subvention
The rationale in brief is that the market model simply does not work for scholarly monographs. My research on this topic supports this rationale. In recent decades, while a few scholarly journal publishers have benefited from increasing revenues and profits, the average sales for scholarly monographs have plummeted. For details, see my dissertation, chapter 6: http://summit.sfu.ca/item/12537#310
My comment: why just the first book? For scholars in fields where monographs are the norm, why not set aside a reasonable amount (e.g. $10,000) for the first-copy cost to publish a monograph every four years or so?