Australian Competitive Grant funding is a major source of university prestige and a significant source of research income. Securing ACG funding has become increasingly competitive and more high-stakes in recent years; particularly since the introduction of the Excellence for Research in Australia (ERA). The inclusion of all academic staff in the ERA count has increased pressure on academics to not only be research-active but to apply for and, if possible, secure ACG funding. In some universities, submitting an ACG application within the first 2 or 3 years of employment is a condition of successfully passing probation and/or achieving promotion. Not surprisingly, while there has been an increase in the funding allocations to the ARC and NHMRC, each funding body has struggled to maintain application success rates against the rise in research funding applications. The ARC has sought to ration available funding by cutting allocations to successful projects, in order to preserve the volume of research to be conducted. However, with some projects receiving as little as 40% of the funding requested and the average receiving only 55%, this strategy has further exacerbated the problem of grants not covering the full costs of conducting the research. The impact on research in the field of education, which “usually gets the crumbs, irrespective of the quality of the work we do” (Goodyear, 2008, p. 4), is high because of the labour intensiveness of educational research, the general lack of respect for qualitative research in the academic research community, and the “messiness” of doing research in schools. This paper seeks to contribute to the conversation about the funding of research by presenting evidence from an ongoing ARC Discovery project that aims to track the experiences of students referred to behaviour schools in NSW. The presentation will outline the re-scoping decisions that became necessary and the challenges involved when attempting to recruit a large enough sample to retain quantitative integrity from the hardest-to-reach parents and young people with just over half of the funds needed to do so. The incidental costs that accrue from complications that occur in the field – which cannot be budgeted for, but would more easily been absorbed under a full cost model of research funding – will be illustrated and discussed through vignettes of conducting research with ghosts who don't like school but who do like lollies, chess and Lego.