The efficiency of a set S of vertices in an undirected graph G=(V,E) i
s defined to be epsilon(S)=\{v\v is-an-element-of V-S and v is adjacen
t to exactly one vertex in S}/, i.e., the number of vertices in V-S th
at are dominated by exactly one vertex in S. The efficiency of a graph
G=(V,E) equals the maximum efficiency of any subset S of vertices of
V. A linear time algorithm is presented for computing the efficiency o
f an arbitrary tree and an NP-completeness proof is given for the prob
lem of deciding if an arbitrary planar bipartite graph has a set S suc
h that epsilon(S) greater-than-or-equal-to k, for some positive intege
r k.