An in-frame deletion (ie deletion of 3 or multiple of 3 nts) should not normally be polar on downstream genes because translation of the gene with the deletion would continue. Polarity arises when you stop translation of the upstream gene and have a long untranslated region.
I worked mostly deletions or insertions on operons during my phd, and in most cases, as Michael J. Benedik said, you should not experience polar effect.
However, in some cases, which can be pretty rare, expression of downstream genes might be altered to a level that gives detectable phenotype. Most operon genes are expressed by translational initiation, but sRNA might have a role in the regulation of the long stretch of mRNA transcribed, which either leads to degradation or enhanced expression.