The American Association of Public Opinion Research has just released a task force report on cell phones and polling. The report is here (pdf). Here are their recommendations:
bq. As can be seen in this report, a great deal of important new research remains to be conducted in the U.S. before telephone survey researchers can conduct RDD surveys of persons reached on their cell phones with the confidence in the findings that is expected by the users of those data. In light of this, there are few recommendations this Task Force believes can be made with confidence at this time. However, as a result of the developments discussed in this report and as an interim step in applying survey methods to cell phones in the U.S., the AAPOR Cell Phone Task Force recommends the following disclosure-related considerations:
bq. All telephone surveys should disclose whether or not the sample includes only landline numbers, only cell phone numbers, or both, and how the numbers were selected from their respective frames.
bq. All RDD telephone surveys with samples that contain cell phone numbers should fully disclose how any weights have been constructed and what population estimates have been used to post-stratify, recognizing that many such parameters are not available at sub-national levels.
bq. RDD telephone surveys targeting subgroups in the U.S. with substantial percentages of adults who live in cell phone only households (e.g., 18 – 29 year olds; renters; and those below the poverty threshold) should sample cell phone numbers or, if this is not feasible, discuss how excluding cell phone numbers may affect the results.
With regard to political polling, two Pew Center studies are key (see here and here). They find little evidence that excluding those who have only a cell phone affects the results of the poll. For an example of how pollsters can screen for cell phone-only households, see here.