I wonder how many of the various governmental surveys (such as employment) or political surveys are dependent on phones as point of contact, and what, if any adjustments are made for no-landlines-slash-cell-phone only respondents? If some percentage of the entire population is cell phone only, how does that skew the surveys results? Some portion of those make the decision to drop their landline logically (that it is a redundant and unnecessary cost) or out of necessity (to reduce their monthly bills). How do those two groups differ? Are they likely to be conservative or liberal? Are the likely to be employed, unemployed, or underemployed? There’s some more data here at Pew Research.