During my Business Research and Quantitative Analysis course for my MBA five years ago, the topic of cellphones in research was a topic that came up many times — a paradigm shift is happening in how we use technology and phones.
While it’s illegal to directly telemarket toward cellphone numbers in most circumstances, political polling is different, and the impacts of being able to use cellphones and landlines shows an enormous gap in American opinions and makeup.
The younger generations are more likely to have cellphones only, the older generations are likely to have landlines only, and those in my own demographic usually have both.
In the business research realm, this is an incredible problem — reaching those who are the most likely to spend all of their disposable incomes is almost impossible when it comes to researching by phone. Where polling in person at malls (and other venues) used to focus almost exclusively on the under-25 crowd, it has expanded because of the inability to reach the under-40 crowd exclusively using cellphones.
But in politics, the difference is profound, and the blend of cellphones and landlines often has an incredible ability to skew the polling data.
In politics ... the blend of cellphones and landlines often has an incredible ability to skew the polling data.
Since 1985, the federally supported Lifeline program subsidized home phone service for those living at 135% of the poverty level. In 2012, a comprehensive overhaul of the system was done, both streamlining the system and including cellphone service for the first time.
Since then, the so-called “Obama-cellphones” have become popular among some of the poorest of our nation, often their only option to connect with other people.
This is creating a trend, where the poorest of the poor are often only reachable through their cellphones, and do not have traditional landlines.
Then you have the younger generations — fascinated with the power and ease of access to the Internet through their cellphones — who have been dumping traditional home phone service.
In the 2016 election, where the split is so profound, the cellphone paradigm is in full play for the Democratic Primary. Sanders claims the lion’s share of the under-45 vote, while Clinton holds an incredible lead with the over-45 vote.
And we are seeing this play out in the polls. In Wisconsin, six major polls have been conducted in the past week. Of those, one was landline only, one was weighted lopsided to landlines (and then the researchers tried to adjust the data to compensate), and three had a blend of cell/landline users that corresponded to the national usage. YouGov/CBS used an opt-in hybrid poll, allowing panelists to join at-will.
And what were the results?
In the three polls where the blend was ‘correct,’ Sanders walks away with a profound lead, with leads above the margin of error in 2 of the 3 polls. In the two polls where cellphones were not properly included, Clinton had the lead, with the leads being greater than the margin of error. The opt-in hybrid poll has Sanders leading, but it is unclear the impact of cellphones on a poll where most of the participants wanted to be talked to.
Can it really be that simple–can cellphones have that much of an impact?
It would seem to be the case in the last week in the Wisconsin polling–and pollsters need to take note.
Polling is never an exact science, but making sure that a study has a representative blend of cell/landline users is essential.