The arbitrary $5000 lower limit for defined equipment purchases in the U.S. results in millions of dollars of lost science funding per year.
The U.S. government defines equipment (stated in 45 CFR Parts 74 and 92), as an article of tangible non-expendable personal property having a useful life of more than one year and an acquisition cost of $5,000 or more per unit. Both the NIH and the NSF use this definition in administering research projects. In grants to the NSF for example, items of needed equipment must be listed individually by description and estimated cost, including tax, and adequately justified. The result: collectively U.S. scientists are losing millions from their research budgets annually.
At most institutions overhead (or indirect cost) is not collected on equipment, while it is on everything else. The national average of overhead costs among universities is 52%. What this means in practice is if a piece of equipment costs $4,999, the effective cost from a grant at an average university is $7,598.48. If the price of the nearly the same item is $1 more, then it costs the grant only $5,000 and saves over $2,500 for the grant. Figure 1 shows all such lost funds (to science) shaded in red. What this means practically if a scientist is trying to decide between two identical pieces of equipment from vendor A, which costs $3,300 and from vendor B, which costs $5,000, the scientist will be a better steward of her research funds if she chooses the latter more expensive tool from vendor B.
Although vendor B’s tool costs significantly more than the tool from vendor A, machine A costs the scientist more on her grant and thus reduces the funds available for materials, supplies, student salaries, etc. This perverted economic incentive (overhead incentivized waste zone is shaded blue in Fig. 1) is obviously a waste of money from the scientific funding agency. It may be tempting to think that such small research expenditures are immaterial as routine scientific grants are in the hundreds of thousands to several million dollars in total expenditures. However, the “little” costs under $5,000 make up for their diminutive stature with volume and represent the majority of non-labor related research costs for standard research projects. It is true that some of these expenditures are for supplies, but much of the core equipment used every day in labs from pipettes and centrifuges to multimeters and soldering irons falls into this <$5,000 category. Thus, the majority of non-labor related expenses costs about 50% more than they should and conservatively results in millions of dollars lost to science annually.
In defense of the standard government science funders, they clearly state that they are willing to be consistent with grantee policies and establish lower limits. Scientists would prefer to have much lower limits set (or simply abolish the limit all together). After all, scientists work hard to earn grant money under extremely competitive circumstances, and thus want to conserve it as much as possible for actual research. This is particularly true of those funded by government organizations like the NSF and DOE where the overhead comes out of the total grant money rather than being added on at the end such as with the NIH. Unfortunately, overhead rates are primarily used to subsidize administrative salaries and building depreciation, neither of which directly benefit research; although it does benefit the administrators that determine university policy. Thus, there is a perverse incentive in place for university administrators to maintain the arbitrary value of $5,000 as a definition of equipment.
This problem is not new. However, now it is becoming even more damaging as this arbitrary accounting value is slowing science in a second way: by effectively creating a tax on the mass diffusion and development of free and open-source hardware (FOSH) for science. The free and open source technological development, which has so powerfully transformed the software world, has been introduced to science. It is now growing rapidly in scientific hardware as generally the costs can be reduced by 90-99%. This is now only possible because of the introduction of FOSH digital manufacturing technology such as RepRap 3-D printers, which enable both new science and greatly reduced difficulty in replicating scientific equipment made by others. There are hundreds of FOSH scientific tools developed throughout the world now, which is sparking a revolution as formerly highly specialized, high-cost scientific equipment are increasingly custom fabricated in-house using digital designs. By using open source hardware design that can be manufactured digitally the relatively minor development costs result in enormous return on investment (ROIs) for scientific funding agencies. With medium-complexity tools where the majority of custom parts can be fabricated with a desktop 3-D printer and assembled in a few hours, ROIs for scientific funders range from 100s to 1,000s of percent after the only a few months of a designs release. This practice is becoming widespread, as any scientific field that adopts this model will immediately gain an innovation advantage not only for equipment, but also in the ability to quickly replicate, verify, and build on one another’s experimental work. For example, scientists have already begun to share optics designs and water testing equipment, that enable rapid diffusion of new experimental methods.
A FOSH model thus makes it possible to stretch research funding much further to accomplish the same or even superior research. Because of the steep discounts for FOSH equipment, the current $5,000 lower limit for equipment effectively penalizes FOSH hardware, which costs less than $50,000 (10% of proprietary costs) to that which costs less than $500,000 (1% of proprietary costs). Thus, the definition of capital equipment is becoming increasingly significant as the current state of development of such FOSH research tools in the few hundred to few thousand dollar range. Following, open source protocols, scientists have been building upon one another’s designs and are creating progressively more sophisticated equipment that is climbing the cost ladder.
In conclusion, the effectiveness of research funding can be improved by eliminating the arbitrary economic minimum from the definition of research equipment. This will effectively recapture all the funds lost to science now (Fig. 1 red area) and remove the incentive to waste research funds by overpaying vendors to meet the arbitrary $5,000 limit (Fig. 1 blue area) on the majority of non-labor research expenditures. More importantly, those who can frugally stretch research dollars the furthest while having the greatest impact by developing FOSH scientific tools will no longer be penalized for their efforts. This should encourage more aggressive and widespread FOSH scientific development. With more of the investments for science making it to research projects, society’s investments will pay larger dividends both directly and indirectly as everyone will have access to high-quality free and open source research equipment at costs below $5,000.