Dominance is not inevitable.

In a diploid organism, a mutant gene that results in elimination of an enzyme activity in the homozygote is almost universally found to be recessive, so that the heterozygote phenotype is virtually indistinguishable from the wild type. It has been argued (H. Kacser & J. A. Burns, Genetics 97, 639-666 (1981)) that there is no need to look to evolution for an explanation of this phenomenon, as it is an inevitable consequence of the low control coefficients for metabolic flux possessed by nearly all enzymes. However, it is possible to envisage pathways in which every enzyme is more than half-saturated, so that moderate changes in the concentration of any enzyme result in substantial changes in metabolic flux. Such behaviour can occur, for example, if the limiting rates of the enzymes decrease as one proceeds along the pathway and the precursor concentration is large compared with the Michaelis constants of all the enzymes. Consequently one does require an explanation in terms of natural selection of why such pathways are apparently not observed in nature.