The high overheads are to be expected in a CRI model, the whole point of which is to return a cash dividend to the shareholder (Government). Having sat for many years on an Australian research funding provider, high overheads were unusual and were usually rejected. Research groups (industry, universities and government research organisations such as CSIRO) were expected to find their "profits" elsewhere. The benefit to them was that research they were interested in could be fully funded (direct costs: salaries, equipment and some travel but not laboratory accommodation or depreciation for example). Not only that, the IP associated with research grants was also closely watched. IP and project milestone and final reports resulting from funding was split with the funder. Reports might be embargoed for commercial reasons but otherwise were considered to be in the public domain (since the public paid for the work). A further incentive was that University funding from the education vote was also linked to performance, the number of PhD's and the number of staff publications had a monetary value to the University so there was an incentive to apply for research that would result in publications in reputable journals, rather than looking on the application as a cash cow for the shareholders.