Most of the authors' analysis rings true, but Dr. Harold Varmus, in particular, contributed enormously to the perverse incentives he now complains about when, as NIH Director, he mandated "modular grants", in which scientists simply request grant support in multiples of $25,000 without the traditional detailed budget and without any salary data. Indeed, scientists were (and still are) expressly forbidden from including in their application any information on exactly how they proposed to spend the requested grant money or what they were paying themselves. The predictable response of the universities (and I speak here from first-hand experience) was to strongly encourage faculty to put larger portions of their salaries onto grants, and be rewarded with higher base salaries. Such policies were enthusiastically sold by department chairmen to upper-level administrators as a way of incentivizing faculty to acquire more grant support, while at the same time raising faculty salaries, all at zero net cost to the University. The fact that all this occurred at the same time as the doubling of the NIH budget only encouraged the process. Now the hard times are here again, money is tight, and support personnel are being let go, but faculty are not giving up their higher salaries and the universities aren't going back to paying faculty from university funds to do research, at least not without a fight.