Animals in Research Facilities / 105The loss of animals in federally funded research facilities rep-
resents a tremendous expense to the American public. More accu-
rately, the expense comes from the replacement of animals. Thus
far, there has been no way to hold facilities accountable. Tax dol-
lars, funneled through the NIH, support research even in facilities
built in high-risk areas, such as fl ood zones. The NIH claims it has
no authority over where a facility is built or how the research is con-
ducted. Animals are replaced without questioning whether they are
needed in the studies because federal policies decide they are not
“really” animals. Because the NIH has been unwilling to hold facil-
ities accountable for disaster-related losses, including all species in
the Animal Welfare Act’s defi nition becomes especially important.
Changing the defi nition of animal in the Animal Welfare Act
would provide greater protection regardless of whether a disaster
strikes. However, in disaster events, this change still leaves labs
holding thousands of animals in impossible situations. The second
step requires taking the Three R’s seriously, especially the man-
date to reduce the numbers of animals used. Currently, research-
ers can employ several loopholes to avoid compliance.^66 Expanding
the coverage of the Animal Welfare Act would eliminate one loop-
hole. At present, because the Animal Welfare Act does not defi ne
the majority of the species used as “animals,” researchers have no
obligation to apply the Three R’s. Simply covering additional spe-
cies will not automatically mean that researchers will take steps
to refi ne, reduce, and replace. For example, although the Animal
Welfare Act requires researchers to consider replacing animal sub-
jects with nonanimal alternatives, confi rmation of this step relies
on an honor system. When researchers submit their proposals for
review by the IACUC, they report having considered nonanimal
alternatives. The committee usually accepts that assertion, because
to do otherwise would be to question (or appear to question) the
design of the experiment.^67 Studies have found that researchers sel-
dom conduct systematic reviews of previous animal experiments
and instead frequently “answer questions that have already been
answered.”^68 Requiring that proposals for new animal experiments