Demand-side management (DSM) critics continue to question the reliabil
ity of DSM program savings and, therefore, the need for funding such p
rograms. In this paper. we examine the issues underlying the discussio
n of reliability of DSM program savings (e.g., bias and precision) and
compare the levels of precision of DSM impact estimates for three uti
lities. Overall, the precision results from all three companies appear
quite similar and. for the most part, demonstrate reasonably good pre
cision levels around DSM savings estimates. We conclude by recommendin
g activities for program managers and evaluators to increase our under
standing of the factors leading to DSM uncertainty and reduce the leve
l of DSM uncertainty.