Hopefully this syllogism will answer your question.

Given the following premises:

  1. In the absence of selection, fitness of individuals are at a theoretical maximum.
  2. If a theoretical maximum fitness is achieved then effective population size is maximum.
  3. If there is an allele that confers both increased *and* decreased fitness you have a genetic conflict (e.g. an allele that turns you green gives you good woodland camouflage but is un-attractive).
  4. If a deleterious allele is linked to a beneficial allele this is similar to having a genetic conflict.

The conclusion follows: 

Strong linkage between beneficial alleles and deleterious alleles (Hill Robertson effects, i.e. 3. and 4.) prevents a theoretically maximum fitness to be attained (1.) and thus a reduced effective population size (2.). Which can only be overcome by high levels of recombination and mating (i.e. by separating deleterious and beneficial alleles).