This paper compares two approaches to estimating the net benefits of a
demand-side management program that subsidizes the purchase of effici
ent equipment. The total resource cost (TRC) test suggests net benefit
s of US$6.9 million. Estimates based on consumer's surplus yield an es
timated net loss of US$6.3 million. This gap is due to differences in
estimates of participants' net benefits. Market failures do not appear
to be significant and thus cannot explain much of the gap. The more p
lausible explanation is that the TRC uses a lower discount rate than p
articipants use for themselves and it omits some costs.