During the last two Presidential election cycles there was little evidence that pollsters omitting cell phones from their samples weren’t getting a true representation of the electorate. A new study by Marist College suggests 2012 may end up being the year that an under sampling or exclusion of cell phone users could have an impact on polls. Marist Institute for Public Opinion Director Lee Miringoff wrote yesterday, “Check out the results from the recent NBC News/Marist Polls of swing states.  In Florida, Romney is +3% with landline voters, but Obama is +23% with cell phones.  In Virginia, it’s pretty much the same… Romney +1% with landlines, Obama is +18% among cells.  Owing to his wider lead in Ohio, Obama is +4% with landline voters but is also +9% with the cell phone electorate.”

According to the January - June 2011 National Health Interview Survey, which monitors “wireless substitution” as part of annual in-person interviews collecting health statistics, in the first six months of 2011 31.6 percent of U.S. households were considered “wireless only.” That’s a significant increase of 5 percent when compared to the number of wireless only households counted in the first six months of 2010. The number of wireless only households has nearly doubled since the first half of 2008.

Nationally, the age demographic changing fastest into wireless only households is the 18-44 group. As a whole 46.3 percent of those households are wireless only - that’s a significant increase of 18.9 percent since 2008. Within that age group the largest growth was in the 18-29 range, where more than half, 52.4 percent, live in wireless only households.

Perhaps not surprisingly, the 45 and up age demographic is only 14.7 percent wireless only, having seen an increase of only 8.7 percent over 2008.

What about Wisconsin? The most recent state-by-state data was collected between July 2009 and June 2010 and according to those statistics Wisconsin had a wireless only household level of 25.3 percent. That’s a jump of 8.9 percent over the numbers collected during the entirety of 2008. Wisconsin is the sixth fastest state in the country when it comes to households converting to wireless only. When compared to the national percentage for that same data range, July 2009 through June 2010 vs. the entirety of 2008, there was a 6.7 increase in  wireless only households in the United States as a whole. Wisconsin outpaced the national average for conversion to wireless only by 2.2 percent. If that pacing continued, these numbers suggest the number of Wisconsin households that were wireless only by the end of 2011 was likely slightly above 32 percent.

That brings us to recent polling in the state of Wisconsin. Few of the recent polls included cell phones and of the ones that did, pollsters don’t break down the number of cell phone users polled by age demographic. For example, the most recent Marquette Law School Poll conducted between May 23 and 26 polls cell phones and landlines. Of those polled, 27 percent said they were reached by cell phone. Marquette’s cell phone polling is likely between 3 to 5 percent below the Wisconsin average.

And, as with the college’s previous poll in mid-May, the age demographics in the most recent poll don’t line up with actual Wisconsin voting trends over the last three elections (two Gubernatorial and one Presidential). On average, those 45 and up made up 54 percent of Wisconsin voters. The number of individuals 45 and up sampled in the most recent Marquette poll was 65 percent, only slightly better than the sample size of 78 percent in the previous poll. The April pre-primary poll had a 45 and up sample size of 59 percent, and in the actual primary election Milwaukee Mayor Tom Barrett outperformed that April poll by 20 points. But the April poll also had a cell phone sample size of 34 percent, much closer to where current trends likely put wireless only households in Wisconsin.

Overall, Tuesday's election will put Marquette's sampling to the test, as Governor Scott Walker and Milwaukee Mayor Tom Barrett face the real Wisconsin voter demographics. Whether cell phone use or age is the stronger indicator of preference, or whether sampling for cell phone use is enough to account for differences across age groups, the results will be a useful benchmark for political polling as households continue to ditch traditional landlines.

Jesse Russell

Co-Editor

Jesse was born and raised in Connecticut, began blogging in 1997, and moved to Madison in 2003. In 2005, he co-founded dane101 along with Kristian Knutson and Shane Wealti. In addition to helping nearly a dozen contributors run this website he's helped launch various events in the city including What's Your Damage?!, the MadPubQuiz of Awesomeness, the Fire Ball Masquerade, Dane101's Freakin' Halloweekend, and more.