Looking for %'s for online vs print serials by subject...

II'm slowly working my way through statistics around our subject funds here at the univ. As budgets are decimated by the big serials deals (disproportionate amount of monies spent on online serials, or even ebooks, most of which are perceived to be STM) our duties to our users, among them specifically leveraging our purchasing power in an equitable way on behalf of our community of users, have become compromised by the political will of our administrators and the big deal. I'm looking for evidence/data that the humanities and social sciences are being gypped, since they are allegedly still heavily print-based. But I'm in search of granular data.

I know I can go through our big deal packages and try and assign titles per subject and percentages per package for each big deal, and then assign for our small packages and one-off titles, but has anyone else done this work? Has anyone done so in pursuit of a different agenda but would be willing to share? Further, has anyone actually done a % breakdown of print vs print+online vs online only by subject such as LC classification? 

I'd like to take pre-existing or and/or newly created data and use it as evidence to force a more equitable distribution of funds, be that based on FTE, Head Count, etc., a breakdown that would respect publishing practices per subject.

Is there anyone out there with this data or knows where it may be found? (and no, I haven't started reviewing LIS databases as yet). And is anyone interested in working on such a research project with me?